Vol. 33, No. 2INFECTION AND IMMUNITY, Aug. 1981, p. 372-3790019-9567/81/080372-08$02.00/0

Distribution of Superoxide Dismutase, Catalase, andPeroxidase Activities Among Treponema pallidum and Other

SpirochetesFAYE E. AUSTIN, JOSEPH T. BARBIERI,t ROBERT E. CORIN,: KATHRYN E. GRIGAS, AND C. D.

COx*Department ofMicrobiology, University ofMassachusetts, Amherst, Massachusetts 01003

Received 27 March 1981/Accepted 8 May 1981

Representative members of Spirochaetales were surveyed for their content ofsuperoxide dismutase (SOD), catalase, and peroxidase activities. Only Leptospiraexhibited peroxidase activity. Obligately anaerobic cultivable Treponema andSpirochaeta possessed no SOD or peroxidative capabilities. Upon polyacrylamidegel electrophoresis, Spirochaeta aurantia, Borrelia hermsi, and five Leptospirabiflexa serovars showed SOD activity associated with one electrophoretic bandwhich was inhibited by H202, suggesting that they were iron-containing dismu-tases. These spirochetes could be distinguished by differences in relative mobilitiesof their SODs. SOD activity, but not catalase activity, was induced aerobically inS. aurantia. All Leptospira interrogans serovars and two L. biflexa serovarslacked significant SOD activity. These SOD-deficient strains of Leptospira, withone exception, possessed high levels of catalase activity. The Nichols strain ofvirulent Treponema pallidum possessed SOD and catalase activities, but lackedperoxidase activity. The SOD in T. pallidum exhibited two electrophoretic bandscontaining copper and zinc, and its relative mobility was identical to that ofpurified rabbit SOD. Immunization of sheep with purified rabbit SOD resulted inantiserum which inhibited both rabbit SOD and T. pallidum SOD assayed byspectrophotometric analysis or activity staining following polyacrylamide gelelectrophoresis. In agarose gel diffusion, precipitin lines of identity were observedbetween purified rabbit SOD and cell extracts of T. pallidum. These dataindicated that the SOD activity detected in T. pallidum was host derived.

Reduction of 02 to water by microorganismsmay result in the production of toxic oxidizedintermediates such as the superoxide radical andH202. Enzymes which scavenge these such assuperoxide dismutase (SOD), catalase, and per-oxidase, may be essential defenses against suchtoxic radicals and compounds. Similarly, 02 me-tabolism is of prime importance in the physiol-ogy and classification of microorganisms. Thespirochetes are excellent examples of microbialdiversity in 02 utilization. Obligately aerobicLeptospira utilize primarily long-chain fattyacids for energy, whereas the anaerobic andfacultatively anaerobic, free-living Spirochaetaare saccharolytic (11). Borrelia hermsi has beencultivated recently and appears to require 02 forgrowth (26). Included in the genus Treponemaare host-associated cultivable anaerobic species(11) as well as pathogenic species, such as Tre-

t Present address: Department of Microbiology, Universityof California, Los Angeles, Los Angeles, CA 90024.

t Present address: Memorial Sloan-Kettering Cancer Cen-ter, New York, NY 10021.

ponema pallidum, which have not been culti-vated in vitro.

Previous investigations in our laboratory in-dicated that T. pallidum was capable of aerobicrespiration (16, 29, 30). These studies revealedthat H202 was produced during the oxidation ofboth pyruvate (3) and reduced nicotinamide ad-enine dinucleotide (30) and that incubation ofwhole cells at high dissolved 02 concentrationsresulted in a loss of pyruvate decarboxylase ac-tivity (4). In addition, reduced flavoproteins andcyanide-resistant respiration, which are knownto be potential sites for superoxide radical gen-eration in several electron transport systems (21,32), have been demonstrated in T. pallidum (29,30). Some means of protection against 02 toxic-ity seemed likely to be required by T. pallidumin vitro, and therefore we have analyzed thedistribution of dismutative and peroxidative en-zymes in this treponeme in comparison withother spirochetes.

MATERIALS AND METHODSOrganisms. Propagation in rabbits, extraction, pu-

372

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

SOD, CATALASE, AND PEROXIDASE IN SPIROCHETES 373

rification, and counting of the Nichols strain of viru-lent T. pallidum have been described previously (16,29, 39). Extraction of treponemes from infected testi-cles was carried out aerobically with shaking for 1 h atroom temperature in phosphate-buffered saline con-taining 2 mM reduced glutathione (pH 7.4). Aftersequential filtration through Nuclepore filters to afinal pore size of 0.8 pm, treponeme suspensions werecentifuged at 17,000 xg for 20 min. High-speed pelletswere rinsed, but not disrupted, in fresh phosphate-buffered saline containing 2 mM reduced glutathioneand recentrifuged at 17,000 x g for 20 mini Pelletswere suspended in a minimal volume of sterile distilledwater, and the supensions were either used immedi-ately or stored in liquid nitrogen. Cultivable Trepo-nema and Spirochaeta were provided by E. Canale-Parola of this department Treponema denticolastrain 10 was cultivated at 35°C in GM-1 medium asdescribed by Blakemore and Canale-Parola (7). Tre-ponema succinifaciens strain 6091 and Treponemabryantii strain RUS-1 were cultivated in RFC mediumcontaining 0.4% glucose intead of cellobiose (41). Spi-rochaeta litoralis strain Ri was cultivated at 30°C ina medium containing 0.4% yeast extract (Difco Labo-ratories, Detroit, Mich.), 0.2% glucose, 0.05% L-CyS-teine hydrochloride, 0.2% Trypticase (BBL Microbi-ology Systems, Cockeysville, Md.), 0.2% peptone(Difco), 50 mM tris(hydroxymethyl)aminomethane-hydrochloride buffer (pH 7.5), and 50% artificial sea-water (5) at a final pH of 7.5. Spirochaeta aurantiastrain Jl was cultivated at 30°C in GTY broth (10)both aerobically on a rotary shaker and anaerobicallyby methods based on those of Hungate (25). B. hermsiwas obtained from R. Kelly and was cultivated asdescribed by him (26). The B and H strains of Lepto-spira biflexa were isolated in this laboratory fromsurface water samples. L. biflexa serovars patoc 1,semaranga, andamana, and ilini 3055 and Lepto-spira interrogans serovars canicola Hond Utrecht,pomona Pomona,pomona Riggs, hardjo Hardjo, grip-potyphosa 1540, and copenhageni M-20 were obtainedorginally from vanous laboratones (23). Cells weregrown in SM-7 medium (14) containing 0.02% Tween80 and 0.2% bovine serum albumin fraction V (MilesLaboratories, Inc., Elkhart, Ind.). Cultivation wasstatic at 30°C, and the cells were harvested when thegrowth achieved a turbidity of 60 to 80 nepholometerunits. The turbidity was measured with a Colemanmodel 9 nephocolorimeter adjusted to 20 nepholome-ter units with a Coleman 81 nepholos standard. CelLswere enumerated with a Petroff-Hausser chamber,and a turbidity of 60 corresponded to approximately1 x 108 to 1.5 x 108 celLs per mL Eswherichia coli Bwas provided by C. B. Thorne of this department andwas grown at 370C on a rotary shaker as described byHassan and Fridovich (22). The medium contained 1%tryptose (Difco), 0.5% peptone (Difco), and 0.1% NaClat a final pH of 7.0.Preparation ofCFE. Cells were harvested by cen-

trifugation, washed in 20 mM potassium phosphatebuffer, (pH 7.4) and suspended in a minimal volumeof this buffer before sonic disruption. Cellular debriswas removed by centrifugation at 20,000 x g for 15min, and the supematant fraction or cell-free extract(CFE) was analyzed. T. palidum CFE was concen-

trated by lyophilization for use in the Ouchterlonyprocedure.Enzyme assays. SOD (EC 1.15.1.1) was assayed as

described by Marklund and Marklund (31) in 300-plreaction mixrs with the rate of autoxidation ofpyrogallol adjusted to 0.01 absorbance unit per min toincrease sensitivity. In addition, SOD was asayed bythe xanthine oxidase-cytochrome c method ofMcCordand Fridovich (33) using 300-d reaction mixtures. Oneunit of SOD activity was defined as the amount ofenzyme required to inhibit the rate of autoxidation ofpyrogallol or the rate of reduction of cytochrome c by50%. Specific activities were expressed as units permilligram of CFE protein. When less than 50% inhi-bition of the rate of autoxidation of pyrogallol wasobserved, the specific activity was calculated from thehighest concentration of CFE protein assayed, whichranged from 0.15 mg of protein for L. interrogansserovar pomona Riggs to 0.42 mg of protein for T.succinifaciens. SOD was separated by electrophoresisof samples in 7.0% polyacrylamide gels (PAGE) with10 mM tis(hydroxymethyl)aminomethane-hydro-chloride-S mM glycine buffer (pH 8.3) as describedby Davis (17), stained for enzyme activity (6), andquantitated by linear nning densitometry. Iron-,manganese-, and copper-zinc-containing enzymes weredisinguished in gels by including 0.1 mM ethylenedi-amine-tetraacetate plus either 5.0 mM H202 (9) or 1.0mM NaCN (45) in the staining reagents.

Catalase (EC 1.11.1.6) was asayed polarographi-cally as described by Rorth and Jensen (38) with fewmodifications. Reactions were conducted at 30°C in3.0 ml of 50 mM potassium phosphate buffer (pH 7.0)saturated with air by sparging. The absolute oxygenconcentration in the buffer was 0.31 umol of 02 per mlas determined by the method of Robinson and Cooper(37). Reactions were initiated by the addition of 10 idof H202 resulting in a final concentration of 15 mM.One unit ofcatalase activity was defined as the amountof enzyme catalyzing the decomposition of 1.0 umol ofH202 per min at 30°C. Specific activities were ex-pressed as units per milligram of CFE proteim Thehighest concentration of CFE protein assayed wasused in calculations of specific activities of less than0.1 U/mg and ranged from 0.22 mg of protein for B.hermsi to 2.86 mg of protein for T. denticola.

Peroxidase (EC 1.11.1.7) was assayed as describedpreviously (46) in 325-pd reaction mixures with o-diisidine as the hydrogen donor. One unit of perox-idase activity was defined as the amount of enzymecatalyzing the decomposition of 1.0 iLmol of H202 permin at 25°C. Specific activities were expressed as unitsper 100 mg of CFE protein. The highest concentrationof CFE protein assayed was used in calculations ofspecific activities of less than 0.1 U/100 mg and rangedfrom 0.05 mg of protein for T. paUidum to 0.66 mg ofprotein for aerobically grown S. aurantia.

Purification of rabbit SOD. SOD was purifiedfrom noninfected rabbit testicles and rabbit erythro-cytes (Pel-Freez, Rogers, Ark.) as described by Mc-Cord and Fridovich (33). Briefly, testicles were groundwith sand and distilled water, or erythrocytes werelysed with distilled water. Hemoglobin was precipi-tated by the addition of cold chloroform and ethanolfollowed by centrifugation. The addition of solid

VoL. 33, 1981

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

374 AUSTIN ET AL.

K2HPO4 to the supernatant fluid resulted in a liquidphase separation. Cold acetone was added to the upperphase containing most of the enzyme activity, andafter centrifugation the precipitate was dissolved in0.05 M potassium phosphate buffer (pH 7.8). TheSOD-containing fraction was purified further by pre-cipitation with (NH4)2SO4 to 90% saturation. Aftercentrifugation the precipitate was dissolved in anddialyzed against 1.0 mM potassium phosphate buffer(pH 7.8) overnight at 40C. This partially purifiedtesticular or erythrocyte fraction represented rabbitSOD. Rabbit SOD contained at least four electropho-retically distinct bands of activity, which are referredto as SOD-1, -2, -3, and -4, in order of increasingmobility. There was no difference in electrophoreticproperties of rabbit testicular SOD and rabbit eryth-rocyte SOD in this system; this finding was in agree-ment with results obtained in previous studies on SODpurified from different tissues (12). The specific activ-ity of rabbit erythrocyte SOD was 3,946 U/mg.

Preparation of rabbit SOD antiserum and itsuse. Preimmune serum was obtained from a sheepbefore starting the immunization procedures. Anti-body to rabbit SOD was produced by intramuscularlyinjecting a sheep with 1.3 mg of rabbit SOD-1 slicedout of polyacrylamide gels. After 3 weeks whole serumwas assayed for antibody by the ring precipitin reac-tion with a clarified crude extract of noninfected rabbittesticles as antigen and decimal dilutions of serum. Notiter of antibody was obtained with undiluted or di-luted test or preimmune sera. Subsequently, a total of7.5 mg of rabbit SOD-1 emulsified in complete Freundadjuvant was administered over a 3-week period. Aring precipitin titer of 1:1,000 was obtained 4 weeksafter the final injection. The immunoglobulin fractionwas prepared by precipitation of serum with 14%Na2SO4 and used in studying its interaction with bothrabbit SOD and T. pallidum CFE by spectrophoto-metric enzyme assays, PAGE, and the Ouchterlonymethod of double diffusion (36). A solution of 0.8%agarose (Miles Laboratories, Inc.), 0.5% NaCl, and0.002% NaN3 was used in the double-diffusion tests.Photographs were taken after at least 20 h of incuba-tion at room temperature. Protein was determined bythe method of Lowry et al. (27) with bovine serumalbumin as a standard.

RESULTSTable 1 summarizes the distribution of SOD,

catalase, and peroxidase activities among severalspirochetes. E. coli B was grown under aerobicconditions and served as a reference in the study.Good agreement in specific activities was ob-served when E. coli CFE was assayed for SODactivity by both the pyrogallol and xanathineoxidase-cytochrome c methods. E. coli catalaseand peroxidase activities were similar to thoseobserved in a previous study (22).Although B. hermsi contained significant

amounts of SOD activity, catalase and peroxi-dase activities were not detected. S. litoralis andthe cultivable treponemes T. denticola, T. suc-cinifaciens, and T. bryantii did not contain ap-

preciable levels ofSOD activity, and these levelsmeasured by spectrophotometric assays weretoo low for detection in polyacrylamide gels.None of these obligately anaerobic spirochetescontained catalase or peroxidase activities, andsimilar negative results had been obtained whenT. bryantii CFE was assayed for catalase andperoxidase activities by a procedure involvingthe oxidation of 3,3'-diaminobenzidine (42). Incontrast, S. aurantia possessed SOD activitywhen cultivated in both aerobic and anaerobicconditions, with a sevenfold increase in activityin the aerobic cultures. The specific activity ofSOD in aerobically grown S. aurantia was com-parable to that observed in aerobically grown E.coli. Although S. aurantia contained no perox-idase activity, catalase was present in aerobicand anaerobic cultures at similar specific activi-ties. These results quantitatively confirm theweakly catalase-positive reaction exhibited byaerobically grown cell pellets when flooded witha solution of H202 (8).A total of 13 serovars of Leptospira, compris-

ing pathogens and water isolates, were includedin this study. All serovars of the pathogenic L.interrogans and B-7 and H-23 serovars of thepresumably nonpathogenic L. biflexa possessedinsignificant SOD activity when measured bytwo spectrophotometric assays and PAGE.However, five other serovars of L. biflexashowed significant levels of SOD activity. Fur-ther attempts were made to demonstrate SODactivity in the L. interrogans serovars and B-7and H-23 serovars of L. biflexa. Cytochrome creductase activity, which might mask SOD ac-tivity in the xanthine oxidase-cytochrome c as-say, was not detected in any of these extracts.To test the possibility that extracts of L. inter-rogans contained an inhibitor of SOD, variousamounts of CFE of several of the L. interrogansserovars were added to CFE of L. biflexa serovarB-16 or semaranga; no inhibition of SOD activ-ity was observed.

Catalase and peroxidase activities were dis-tributed among the Leptospira serovars in apattern which had been observed previously insemiquantitative assays performed in this labo-ratory (14). Peroxidase activities fluctuated ap-proximately 30-fold within the genus, with thepathogenic L. interrogans serovars possessinglower amounts than the L. biflexa serovars. Awide range of catalase values was observed, withthe L. interrogans serovars possessing higheractivities than the L. biflexa serovars. Therewere only two exceptions, L. biflexa serovars H-23 and illini, which contained relatively highlevels of catalase.

B. hermsi, S. aurantia, and certain L. biflexaserovars (Table 1) exhibited a single SOD activ-

INFECT. IMMUN.

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

SOD, CATALASE, AND PEROXIDASE IN SPIROCHETES 375

TABLE 1. Distribution ofSOD, catalase, and peroxidase activities among spirochetesSOD

Electrophoretic bands Catalase (U/ Peroxidase

U/mg' Relative Inhibition by: mg)b (U/100 mg)cmobility H202 NaCN

T. pallidum 20 (17) 0.33 + + 1.66 <0.110.38 + +

T. denticola <3 od <0.01 <0.02T. succinifaciens <3 (2) 0 <0.01 <0.02T. bryantii <3 0 <0.02 <0.02B. hermsi 35 0.54 + - <0.09 <0.06S. litoralis <3 (2) 0 <0.01 <0.02S. aurantia (anaerobic) 30 0.24 + - 0.22 <0.02S. aurantia (aerobic) 220 0.24 + - 0.21 <0.01L. biflexa serovars

B-16 75 (51) 0.40 + - 0.19 10.2patoc 54 (42) 0.39 + - 0.34 10.2semaranga 56 (61) 0.39 + - 0.34 6.5andamana 70 (72) 0.40 + - 0.52 19.7illini 60 (62) 0.40 + - 37.20 3.6H-23 5 (6) 0 6.42 2.9B-7 3 (7) 0 0.42 6.3

L. interrogans serovarscanicola <5 (3) 0 245.4 0.6pomona Pomona <6 (1) 0 103.8 1.2pomona Riggs <6 (4) 0 488.2 0.9hardjo <4 (3) 0 60.8 1.4grippo4 yphosa <5 (4) 0 14.2 1.0copenhageni <6 (3) 0 70.2 1.4

E. coli B (aerobic) 245 (250) 0.31 - - 49.0 5.20.44 + -0.60 + -

aJUnits per milligram of CFE protein measured by the pyrogallol method. Values in parentheses representunits per milligram of CFE protein measured by the xanthine oxidase-cytochrome c method. One unit is theamount of enzyme required to inhibit the rate of autoxidation of pyrogallol or the rate of reduction ofcytochrome c by 50%.

b Units per milligram of CFE protein. One unit is the amount of enzyme catalyzing the decomposition of 1.0Amol of H202 per min at 300C.

'Units per 100 mg of CFE protein. One unit is the amount of enzyme catalyzing the decomposition of 1.0Mmol of H202 per min at 250C.

d No activity band observed.

ity band which was sensitive to only H2O2, indi-cating that theywere iron-containing dismutases(9); these genera could be distinguished on thebasis of differences in the relative mobilities oftheir SODs. PAGE of E. coli CFE resulted inthree bands ofSOD activity, with the two faster-migrating bands inactivated by H202. The slow-est migrating band in E. coli was insensitive toH202 and NaCN, suggesting that it was a man-

ganese-containing dismutase (9).In contrast, virulent T. pallidun possessed

SOD and catalase activities but lacked peroxi-dase activity at the highest protein concentra-tion tested. This treponeme had at least twoelectrophoretic bands of SOD activity that wereinactivated by both H202 and NaCN, a charac-teristic of copper-zinc-containing dismutases

(45). The indication that T. pallidum possesseda copper-zinc-containing SOD appeared to beunique since that enzyme is associated primarilywith eucaryotic cells (45). However, the trepo-nemes used in this study had been extractedfrom rabbit testicles, and Alderete and Baseman(1) had demonstrated the presence of a numberof host proteins on the surface of T. pallidum.Based upon these observations, the possibilitythat the copper-zinc-containing SOD in T. pal-lidum originated in host tissue was further ex-plored.

Partially purified rabbit SOD was subjectedto PAGE and displayed four distinct bands ofactivity. These multiple bands are characteristicof purified eucaryotic SOD and are consideredto be either products of the purification process

VOL. 33, 1.981

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

376 AUSTIN ET AL.

(40) or isozymes (45). Relative mobilities of rab-bit SOD-1, -2, -3, and -4 were 0.33, 0.38, 0.44, and0.48, respectively, and all bands were sensitiveto H202 and NaCN. Rabbit SOD-1 and -2 wereidentical to the two SOD bands detected in T.pallidum CFE in terms of relative mobility andsensitivity to both H202 and NaCN.The reactivity of sheep anti-rabbit SOD-1 se-

rum with homologous rabbit SOD was deter-mined spectrophotometrically and, after resolu-tion, by PAGE by the inhibition ofSOD activity.A diluted rabbit SOD fraction containing ap-proximately 1 U of SOD per mg of protein wasincubated with equal volumes of twofold serialdilutions of antiserum overnight at 4°C. Afterincubation immune complexes were detected asprecipitates in the mixtures containing lowerdilutions of antiserum and were removed bycentrifugation. Analysis of the supernatant frac-tions by the pyrogallol method revealed thatdilutions of 1:8 or less of antiserum completelyinhibited the SOD activity. Upon PAGE of su-pernatant fractions followed by activity stainingfor SOD, no enzymatic activity was observedwith a 1:2 dilution ofantiserum in the area wherefour bands appeared using nonabsorbed rabbitSOD. These results indicated that antibodyraised against rabbit SOD-1 reacted with all fourrabbit SOD PAGE bands, thereby eliminatingenzymatic activity and electrophoretic bands.

Similarly, sheep anti-rabbit SOD-1 serum alsoreacted with T. pallidum SOD, resulting in thedisappearance of the two PAGE bands of tre-poneme SOD. Additional support for the pres-ence of rabbit SOD in T. pallidum CFE wasobtained by agarose gel double-diffusion withsheep anti-rabbit SOD-1 serum (Fig. 1). RabbitSOD yielded three precipitin arcs a, b, and c,

FIG. 1. Agarose gel double-diffusion with T. pal-lidum CFE and purified rabbit SOD. Well 1, 180 pgof T. pallidum CFE protein containing 3.0 U ofSODactivity; well 2, 30 pg ofpurified rabbit SOD contain-ing 117 U of enzyme activity; well 3, 10 ,ul of sheepanti-rabbit SOD-I serum.

with arc b being very sharp and arcs a and cbeing diffuse. T. pallidum CFE exhibited twoprecipitin arcs which showed complete identitywith the corresponding arcs b and c of rabbitSOD. The absence of arc a in T. pallidum CFEmay have been due to the sensitivity of the assayand the limited amounts of treponeme proteincurrently available. The agarose gel was stainedfor SOD activity by the method employed forstaining polyacrylamide gels (6), and the resultsare shown in Fig. 2. Achromatic zones, indicativeofSOD activity, extended from the antigen wellsin all directions. However, the precipitin arcs ofboth rabbit SOD and T.pallidum CFE appearedblue, which indicated the absence of enzymeactivity in the arcs and thus confirmed the abil-ity of the antibody to inhibit rabbit SOD activ-ity.

DISCUSSIONSOD has been regarded in recent years as an

enzyme ubiquitous among 02-metabolizing or-ganisms and absent in obligately anaerobic bac-teria. From these observations McCord et al.(34) proposed an enzyme-based theory of obli-gate anaerobiosis which implicated SOD in pro-tection against 02 toxicity. Indeed, the absenceof SOD, as well as catalase and peroxidase, in T.denticola, T. succinifaciens, T. bryantii, and S.litoralis in the present study would seem tosupport that concept. However, subsequentstudies have shown that various anaerobic bac-teria contain SOD activity (19, 24, 43), and re-cently it has been reported that some strains ofthe obligately aerobic Neisseria gonorrhoeae(35) and Mycoplasma pneumoniae (28), as wellas the facultative Lactobacillus plantarum (2),lack detectable levels of SOD activity. Our find-

FIG. 2. Agarose gel double-diffusion with T. pal-lidum CFE and purified rabbit SOD. Well contentsare the same as in Fig. 1. The agarose gel was stainedfor SOD activity as described by Beauchamp andFridovich (6).

INFECTr. IMMUN.

I

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

SOD, CATALASE, AND PEROXIDASE IN SPIROCHETES 377

ing that B-7 and H-23 serovars of L. biflexa andall serovars of L. interrogans tested lacked de-tectable activities of SOD is another exceptionto the presumed requirement ofSOD for aerobiclife.The SOD-deficient strains of Leptospira, with

the exception of L. biflexa B-7, resemble theSOD-deficient strains of N. gonorrhoeae (35) inthat they possess extremely high levels of cata-lase activity. L. biflexa B-7 and other serovarsof L. biflexa possess peroxidase activity. Thesefindings generally support a previous semiquan-titative survey of peroxidative activities in Lep-tospira (14). The presence of catalase or perox-idase activities (or both) may alleviate the needfor SOD in these spirochetes. Kinetic data haveindicated that the peroxidase of L. biflexa func-tions optimally at low concentrations of H202,whereas the catalase of L. interrogans functionsoptimally at higher concentrations of H202 (15).Therefore, H202 may be scavenged efficiently bythese enzymes so that it cannot react with metalsreduced by the superoxide radical to form themore reactive hydroxyl radical (20). The pres-ence of SOD in some serovars of L. biflexa maybe indicative of excessive formation of such toxicoxidized intermediates.

Diversity in the ability of spirochetes to scav-enge oxidized intermediates is exemplified fur-ther by the distribution of peroxidative enzymesin B. hermsi and S. aurantia. B. hermsi requiresthe presence of 02 for growth in vitro (26).However, the amount of 02 must be reduced,which suggests that this spirochete possessesonly limited defenses against toxic oxidized in-termediates. This is consistent with our findingsthat B. hermsi possessed no detectable amountof catalase or peroxidase, but did contain an ironSOD. S. aurantia is a free-living, facultativelyanaerobic spirochete which was cultivated underboth aerobic and anaerobic conditions in thisstudy. The level of catalase in this spirochetewas relatively low and appeared to be nonindu-cible by 02. However, the SOD in S. aurantiawas induced by 02 with about a sevenfold in-crease in activity during aerobic growth. Anincreased rate of SOD synthesis in response to02 has been observed in E. coli and Streptococ-cus faecalis (18), but S. aurantia differs fromthese bacteria in that its inducible SOD con-tained iron rather than manganese.The evidence permits us to conclude that SOD

from T. pallidum originated in rabbit tissue,which is consistent with the finding of Aldereteand Baseman (1) that a surface coat of hostproteins exists on the treponeme. The observedcatalase activity may represent either an addi-tional host-derived protein or an intrinsic tre-poneme enzyme. Turner and Hollander (44) sug-

gested that the development of syphilitic infec-tion may not depend on the animal species butrather on host conditions of temperature, 02levels, hormonal influences, metals, etc. Perhapsa dependence on host SOD, and possibly hostcatalase, by T. pallidum for growth in oxygen-ated tissues could explain the reported differ-ences among animals in susceptibility to infec-tion by this particular strain. It is attractive tospeculate that human or rabbit SOD and cata-lase would function better for T. pallidumfreshly extracted from humans or rabbits thanenzymes derived from more resistant animals.In our assay conditions T. pallidum resembles

the host-associated cultivable treponemes inlacking their own dismutative and peroxidativeenzymes (Table 1). Although the rabbit SODassociated with T. pallidum is enzymaticallyactive, we cannot rule out the possibility that itspresence in T. pallidum may be fortuitous. How-ever, one must consider the possibility that rab-bit SOD, as well as catalase, does function forthe benefit of T. pallidum. Resolution of thecellular location of rabbit SOD would be essen-tial in evaluating the functional role of this en-zyme. It has been proposed that the surface coatof host proteins may protect T. pallidum bymasking immunogens or promoting membraneintegrity (1, 13). Thus, rabbit SOD associatedwith the outermost surface of the treponememight prevent cellular damage by catalyzing thedismutation of superoxide radicals generated ex-tracellularly or by shielding key outer membranecomponents from oxidation or by both. Theenzyme might also catalyze the dismutation ofany intracellular superoxide radicals which mayhave diffused to the exterior of the cell. It isdifficult to envision such diffusion occurringwithout concomitant cellular damage; however,one must remember that this treponeme has notyet been cultivated in vitro.The numerous host proteins that have been

detected previously were either loosely or avidlyassociated with the treponeme surface, and thepresence of noncompetitive binding sites wassuggested (1). The mechanism of binding of rab-bit SOD to T. pallidum is as yet unknown. Ifrabbit SOD were loosely adsorbed to the trepo-neme surface, then it might readily dissociateupon extraction of cells from tissue or duringmaintenance in vitro. The level of rabbit SODdetected may only be a fraction of that requiredby the treponeme for protection against oxidizedintermediates of 02 when removed from tissue.The inhibiting antibody prepared in this studymay be useful in monitoring the association ofrabbit SOD with the treponeme and furtherelucidating its role in the aerobic metabolism ofT. pallidum.

VOL. 33, 1981

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

378 AUSTIN ET AL.

ACKNOWLEDGMENTSThis investigation was supported by Public Health Service

grant AI-10982 from the National Institute of Allergy andInfectious Diseases.We thank E. Canale-Parola and co-workers for supplying

the cultivable anaerobic Treponema and Spirochaeta.

LITERATURE CITED

1. Alderete, J. F., and J. B. Baseman. 1979. Surface-associated host proteins on virulent Treponema palli-

dum. Infect. Immun. 26:1048-1056.2. Archibald, F. S., and I. Fridovich. 1981. Manganese

and defenses against oxygen toxicity in Lactobacillusplantarum. J. Bacteriol. 145:442-451.

3. Barbieri, J. T., and C. D. Cox. 1979. Pyruvate oxidationby Treponema pallidum. Infect. Immun. 25:157-163.

4. Barbieri, J. T., and C. D. Cox. 1981. Influence of oxygenon respiration and glucose catabolism by Treponema

pallidum. Infect. Immun. 31:992-997.5. Baumann, P., L. Baumann, and M. Mandel. 1971.

Taxonomy of marine bacteria: the genus Beneckea. J.Bacteriol. 107:268-294.

6. Beauchamp, C., and L. Fridovich. 1971. Superoxidedismutase: improved assays and assay applicable toacrylamide gels. Anal. Biochem. 44:276-287.

7. Blakemore, R. P., and E. Canale-Parola. 1976. Argi-nine catabolism by Treponema denticola. J. Bacteriol.128:616-622.

8. Breznak, J. A., and E. Canale-Parola. 1975. Morphol-ogy and physiology of Spirochaeta aurantia strainsisolated from aquatic habitats. Arch. Microbiol. 105:1-12.

9. Britton, L., D. P. Malinowski, and I. Fridovich. 1978.Superoxide dismutase and oxygen metabolism in Strep-tococcus faecalis and comparisons with other orga-

nimms. J. Bacteriol. 134:229-236.10. Canale-Parola, E. 1973. Isolation, growth and mainte-

nance of anaerobic free-living spirochetes, p. 61-73. InJ. R. Norris and D. W. Ribbons (ed.), Methods inmicrobiology, vol. 8. Academic Press, Inc., New York.

11. Canale-Parola, E. 1977. Physiology and evolution ofspirochetes. Bacteriol. Rev. 41:181-204.

12. Carrico, R. J., and H. F. Deutsch. 1969. Isolation ofhuman hepatocuprein and cerebrocuprein: their iden-tity with erythrocuprein. J. Biol. Chem. 244:6087-6093.

13. Christiansen, S. 1963. Protective layer covering patho-genic treponemata. Lancet i:423-425.

14. Corin, R. E., E. Bogg8, and C. D. Cox. 1978. Enzymaticdegradation of H202 by Leptospira. Infect. Immun. 22:672-675.

15. Corin, R. E., and C. D. Con. 1980. Characterization ofleptospiral catalase and peroxidase. Can. J. Microbiol.26:121-129.

16. Cox, C. D., and M. K. Barber. 1974. Oxygen uptake byTreponema pallidum. Infect. Immun. 10:123-127.

17. Davis, B. J. 1964. Disc gel electrophoresis. II. Methodand application to human serum proteins. Ann. N.Y.Acad. Sci. 121:404427.

18. Gregory, E. M., and I. Fridovich. 1973. Induction ofsuperoxide dismutase by molecular oxygen. J. Bacteriol.114:543-548.

19. Gregory, E. M., W. E. C. Moore, and L. V. Holdeman.1978. Superoxide dismutase in anaerobes: survey. Appl.Environ. Microbiol. 35:988-991.

20. Halliwell, B. 1978. Superoxide-dependent formation ofhydroxyl radicals in the presence of iron chelates. FEBSLett. 92:321-326.

21. Hassan, H. M., and I. Fridovich. 1977. Regulation ofthe synthesis of superoxide dismutase in Escherichiacoli: induction by methyl viologen. J. Biol. Chem. 252:7667-7672.

22. Hassan, H. M., and L. Fridovich. 1978. Regulation ofthe synthesis of catalase and peroxidase in Escherichiacoli. J. Biol. Chem. 253:6445-6450.

23. Henneberry, R. C., and C. D. Cox. 1968. Antigenicanalysis of water forms of Leptospira. J. Bacteriol. 96:1419-1420.

24. Hewitt, J., and J. G. Morris. 1975. Superoxide dismu-tase in some obligately anaerobic bacteria. FEBS Lett.50:315-318.

25. Hungate, R. E. 1969. A roll tube method for cultivationof strict anaerobes, p. 117-132. In J. R. Norris and D.W. Ribbons (ed.), Methods in microbiology, vol. 3B.Academic Press, Inc., New York.

26. Kelly, R. 1973. Cultivation of Borrelia hermsi. Science173:443-444.

27. Lowry, 0. H., N. J. Rosebrough, A. L Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

28. Lynch, R. E., and B. C. Cole. 1980. Mycoplasma pneu-moniae: a prokaryote which consumes oxygen and gen-erates superoxide but which lacks superoxide dismu-tase. Biochem. Biophys. Res. Commun. 96:98-105.

29. Lysko, P. G., and C. D. Cox. 1977. Terminal electrontransport in Treponema pallidum. Infect. Immun. 16:885-890.

30. Lysko, P. G., and C. D. Cox. 1978. Respiration andoxidative phosphorylation in Treponema pallidum. In-fect. Immun. 21:462-473.

31. Marklund, S., and G. Marklund. 1974. Involvement ofthe superoxide anion radical in the autoxidation ofpyrogallol and a convenient assay for superoxide dis-mutase. Eur. J. Biochem. 47:469-474.

32. Massey, V., S. Strickland, S. G. Mayhew, L G. How-ell, P. C. Engel, R. G. Matthews, M. Schuman, andP. A. Sullivan. 1969. The production of superoxideanion radicals in the reaction of reduced flavins andflavoproteins with molecular 02. Biochem. Biophys.Res. Commun. 36:891497.

33. McCord, J. M., and I. Fridovich. 1969. Superoxidedismutase: an enzymic function for erythrocuprein(hemocuprein). J. Biol. Chem. 244:6049-6055.

34. McCord, J. M., B. B. Keele, Jr., and L. Fridovich. 1971.An enzyme-based theory of obligate anaerobiosis: thephysiological function of superoxide dismutase. Proc.Natl. Acad. Sci. U.S.A. 68:1024-1027.

35. Norrod, P., and S. A. Morse. 1979. Absence of super-oxide dismutase in some strains of Neisseria gonor-rhoeae. Biochem. Biophys. Res. Commun. 90:1287-1294.

36. Ouchterlony, 0. 1968. Handbook of immunodiffusionand immunoelectrophoresis, p. 2147. Ann Arbor Sci-ence Publishers, Inc., Ann Arbor, Michigan.

37. Robinson, J., and J. M. Cooper. 1970. Method of deter-mining oxygen concentrations in biological media, suit-able for calibration of the oxygen electrode. Anal. Bio-chem. 33:390-399.

38. Rorth, M., and P. K. Jensen. 1967. Determination ofcatalase activity by means of the Clark oxygen elec-trode. Biochim. Biophys. Acta 139:171-173.

39. Schiller, N. L, and C. D. Cox. 1977. Catabolism ofglucose and fatty acids by virulent Treponema paii-dum. Infect. Immun. 16:60-68.

40. Stansell, M. J., and H. F. Deutsch. 1965. Physicochem-ical studies of crystalline human erythrocuprein. J. Biol.Chem. 240:4306-4311.

41. Stanton, T. B., and E. Canale-Parola. 1979. Enumera-tion and selective isolation of rumen spirochetes. Appl.Environ. Microbiol. 38:965-973.

42. Stanton, T. B., and E. Canale-Parola. 1980. Treponemabryantii sp. nov., a rumen spirochete that interacts withcellulolytic bacteria. Arch. Microbiol. 127:145-156.

43. Tally, F. P., B. R. Goldin, N. V. Jacobus, and S. L.

INFECT. IMMUN.

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

VOL. 33, 1981 SOD, CATALASE, AND PEROXIDASE IN SPIROCHETES 379

Gorbach. 1977. Superoxide dismutase in anaerobic 45. Weisiger, R. A., and I, F'ridovich. 1973. Superoxidebacteria of clinical significance. Infect. Immun. 16:20- dismutase: organelle specificity. J. Biol. Chem. 248:25. 3582-3592.

44. Turner, T. B., and D. H. Hollander. 1957. Biology of 46. Worthington Biochemical Corporation. 1972. Worth-the treponematoses. Monograph series no. 35. World ington enzyme manual, p. 43-45. Worthington Biochem-Health Organization, Geneva. ical Corp., Freehold, N.J.

on May 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from