Bactericidal and Cytotoxic Effects of Hypothiocyanite-Hydrogen Peroxide Mixtures

Transcription

1 INFECTION AND IMMUNITY, June 1984, p /84/6581-6$2./ Copyright 1984, American Society for Microbiology Vol. 44, No. 3 Bactericidal and Cytotoxic Effects of Hypothiocyanite-Hydrogen Peroxide Mixtures JAN CARLSSON,'* MAY-BRITT K. EDLUND1, AND LENNART HANSTROM2 Departments of Oral Microbiology1 and Periodontology,2 University of Umea, S-9187 Umea, Sweden Received 14 December 1983/Aepted 24 February 1984 Lactoperoxidase catalyes the oxidation of thiocyanate by hydrogen peroxide into hypothiocyanite, a reaction which can protect bacterial and mammalian cells from killing by hydrogen peroxide. The present study demonstrates, however, that lactoperoxidase in the presence of thiocyanate can actually potentiate the bactericidal and cytotoxic effects of hydrogen peroxide under specific conditions, such as when hydrogen peroxide is present in the reaction mixtures in excess of thiocyanate. The toxic agent was also formed in the absence of lactoperoxidase in a reaction between hypothiocyanite and hydrogen peroxide. Sulfate, sulfite, cyanate, carbonate, and ammonia, which have been postulated to be formed in the chemical oxidation of hypothiocyanite by hydrogen peroxide, were not bactericidal and did not potentiate the bactericidal effect of hydrogen peroxide. Cyanosulfurous acid, the only other postulated product of the chemical oxidation of hypothiocyanite by hydrogen peroxide, may be the killing agent. Lactoperoxidase catalyses the oxidation of thiocyanate by hydrogen peroxide into hypothiocyanite (OSCN-) (4, 19). Lactoperoxidase and thiocyanate are components of the human salivary secretions (25), and hydrogen peroxide is excreted by oral bacteria (2). The OSCN- ion inhibits bacterial glyceraldehyde 3-P dehydrogenases (11, 24) and thereby stops the bacterial production of acids from sugars. The inhibition of bacterial acid production by OSCN- has been implicated as playing an important role in the prevention of dental caries (17). Hydrogen peroxide can be highly toxic to mammalian cells (3, 13, 14) and bacteria (2, 1). This effect of hydrogen peroxide is alleviated, however, in the presence of lactoperoxidase and thiocyanate (2, 8, 13). It has therefore been suggested that lactoperoxidase and thiocyanate of the salivary secretions play an important role in protecting the epithelial cells of the oral mucous membranes against the hydrogen peroxide excreted by bacteria coloniing the oral surfaces (2, 13). In this role, lactoperoxidase in the presence of thiocyanate detoxifies hydrogen peroxide by converting it into OSCN-, and OSCN- prevents bacteria from excreting hydrogen peroxide by inhibiting glyceraldehyde 3-P dehydrogenase. Because of this inhibition, no NADH is generated in the bacteria, and the hydrogen peroxide-producing NADH oxidases become short of their substrate, NADH (11). This inhibition of glycolysis usually has a bacteriostatic effect. In recent studies, significant levels of OSCN- have been found in saliva collected directly from the ducts of the salivary glands (22, 29). This indicates that hydrogen peroxide is actually produced within the salivary glands; thus, lactoperoxidase and thiocyanate may also play an important role in protecting the salivary glands and ducts against hydrogen peroxide toxicity. The products of the lactoperoxidase-thiocyanate-hydrogen peroxide reaction have also been reported to be bactericidal (6, 7, 33, 36, 37). This effect has been ascribed to OSCN-, but it has also been suggested that higher oxyacids * Corresponding author. 581 of the thiocyanate ion, cyanosulfurous and cyanosulfuric acids, may be formed in the lactoperoxidase reaction, and these acids may be the effective molecular species in the killing (6, 15, 31). We now report that a mixture of OSCN- and hydrogen peroxide in the absence of lactoperoxidase was more bactericidal and more toxic to mammalian cells than was hydrogen peroxide alone. MATERIALS AND METHODS Microorganisms. Peptostreptocous anaerobius VPI 433 (ATCC 27337) was kept on blood agar plates, and Escherichia coli K-12 K37, an Str/r mutant strain of W 312 (5), was kept on a minimal agar medium at 4 C in an anaerobic box with an atmosphere of 1% hydrogen in nitrogen. Mammalian cells. HeLa cells were kindly provided by E. Lundgren, University of Umea, Umea, Sweden. The cells were cultured in an atmosphere of 5% carbon dioxide in air in Eagle minimal essential medium containing 1% fetal calf serum, penicillin (6 mg liter-1 streptomycin (1 mg liter-'), and gentamicin (2 mg liter-'). Media. Peptone-yeast extract-glucose broth was prepared as described by Holdeman et al. (16), but instead of being prereduced and saturated with carbon dioxide before being autoclaved, the broth was autoclaved and then stored in the anaerobic box for at least 2 days before it was used. The MOPS (3-N-morpholino propanesulfonic acid) buffer solution was a modification of a medium devised by Neidhardt et al. (26) and contained 4 mm MOPS, 4 mm tricine,.3 mm potassium sulfate,.4,um calcium chloride,.5 mm magnesium chloride, and 5 mm sodium chloride. This solution was adjusted to ph 7.4 by KOH and sterilied by filtration. The phosphate-buffered saline contained 137 mm sodium chloride, 2.7 mm potassium chloride, 1.5 mm monopotassium phosphate, and 7.7 mm disodium phosphate (ph 7.3). A minimal medium contained the MOPS buffer solution, 2 mm glucose, 1 mm ammonium bicarbonate, 1 mm dipotassium phosphate, and.2 mm ferric chloride. The minimal agar medium was supplemented with 16 g of agar (Difco Laboratories, Detroit, Mich.) per liter. Blood agar medium was prepared as described by Holdeman et al. (16). The

2 582 CARLSSON, EDLUND, AND HANSTROM defibrinated horse blood in the medium was hemolyed by freee-thawing. Evaluation of cytotoxicity and bactericidal effects. The mammalian cells were cultured in 25-cm2 plastic culture flasks until the growth was confluent. They were detached with.1% trypsin-.2% EDTA in phosphate-buffered saline, washed in phosphate-buffered saline, and suspended in the culture medium; portions of the suspension were added to the culture medium of 25-cm2 culture flasks to obtain a final volume of 5 ml and a cell density of 1 cells per cm2. The number of cells attached to the bottom surface of the flask was counted with an inverted phase-contrast microscope supplied with a grid with squares (1 by 1 mm). After 4 h, ca. 7% of the cells had attached to the bottom surface. After 2 h, the cells were counted, the medium was decanted, and 5 ml of phosphate-buffered saline (at 22 C) containing various combinations of hydrogen peroxide, thiocyanate, lactoperoxidase, and a lactoperoxidase-free OSCN- preparation was added. The solution was decanted after 3 min, and the cells were exposed to fresh test solution for another 3 min. The test solution was then replaced by culture medium. The numbers of cells attached to the bottom surface of the flasks were counted after 2 and 3 days. Triplicate flasks were run in all experiments. P. anaerobius was grown at 37 C in peptone-yeast extractglucose broth, and E. coli was grown in the minimal medium under anaerobic conditions. When the cultures were in exponential growth phase and had a density of 18 cells ml-,, they were diluted in MOPS buffer solution or in phosphate-buffered saline to a density of ca. 1 x 14 to 1.5 x 14 cells ml-'. Within 2 min after the start of the dilution procedure, a.2-ml sample of this suspension was added to.8 ml of a reaction mixture containing MOPS buffer solution or phosphate-buffered saline with various additions. In most experiments hydrogen peroxide was added to the reaction mixture after the bacteria. Samples (.1 ml) were taken at various times from the reaction mixtures and spread over the surface of duplicate blood agar plates. The plates were incubated for at least 2 days under anaerobic conditions, and the numbers of surviving organisms were determined. Preparation of lactoperoxidase-free OSCN- solution. Each stirred 1-ml ultrafiltration cell (Amicon Corp., Lexington, Mass.) fitted with a Diaflo membrane (PM 3) contained 9 ml of MOPS buffer solution or phosphate-buffered saline,.1 or 1 mm KSCN and lactoperoxidase (1 jig ml-'). Hydrogen peroxide was added to this reaction mixture to give a final concentration of.1 or 1 mm, and 5 min after this addition the solution was filtered. The concentration of OSCN- in the filtered solution was ca. 4 or 4 jim, respectively, as determined with the 2-nitro-5-thiobenoic acid reagent. This reagent was prepared as previously described (11), and the concentration of 2-nitro-5-thiobenoic acid was calculated assuming an extinction coefficient of 14,13 M-1cm-1 at 412 nm (34). Before the solution containing OSCN- was used in the reaction mixture, it was passed through a.22-p.m filter (Millipore Corp., Bedford, Mass.). Chemicals. Hydrogen peroxide (3%[wt/wt]; Perhydrol) was from E. Merck AG, Darmstadt, Federal Republic of Germany; lactoperoxidase (from milk), tricine, and MOPS were from Sigma Chemical Co., St. Louis, Mo.; potassium thiocyanate was from Riedel De Haen AG, Seele-Hannover, Federal Republic of Germany; streptomycin sulfate was from Glaxo Laboratories Ltd., England; benyl penicillin sodium salt were from Astra, Sodertalje, Sweden; gentamicin was from Flow Laboratories, Svenska AB, Stockholm, Sweden; and 1,4-diaa-bicyclo(2.2.2)octane (DABCO) was from EGA-Chemie Gesellschaft mbh & Co., Steinheim, Federal Republic of Germany. RESULTS Hydrogen peroxide (8,uM) killed 9% of the cells of P. anaerobius within ca. 4 min (Fig. 1, line 2). Lactoperoxidase in the presence of 1,uM thiocyanate protected the cells from the bactericidal effect of hydrogen peroxide (Fig. 1, line 3), whereas lactoperoxidase potentiated the bactericidal effect of hydrogen peroxide in the absence of thiocyanate (Fig. 1, line 4). In a mixture of lactoperoxidase, 1,uM KSCN, and 8,uM hydrogen peroxide in MOPS buffer solution, ca. 4,uM OSCN- was formed. P. anaerobius survived in this solution after lactoperoxidase had been removed from it (Fig. 1, line 5). When P. anaerobius was stored in this lactoperoxidasefree solution containing 32,uM OSCN- for 5 min and then exposed to 8,uM hydrogen peroxide, the toxicity of hydrogen peroxide was highly potentiated (Fig. 1, line 6). When cells of P. anaerobius were exposed to 8 pum hydrogen peroxide in the presence of lactoperoxidase and various concentrations of KSCN, only concentrations higher than 8 jim KSCN fully protected the cells from killing by u Lc CO).1 INFECT. IMMUN TIME (min) FIG. 1. Killing of P. anaerobius after exposure to various combinations of lactoperoxidase-thiocyanate-hydrogen peroxide, showing the surviving fraction of cells, under the following conditions: after storage for 1 h in MOPS buffer solution (line 1); in the presence of 8,uM hydrogen peroxide (line 2); in the presence of lactoperoxidase (1,ug ml-'), 8 F.M hydrogen peroxide, and.1 mm KSCN (line 3); in the presence of lactoperoxidase (1 jig ml-') and 8,uM hydrogen peroxide (line 4); after storage for 1 h in a 32 I.M lactoperoxidasefree OSCN- preparation (line 5); after storage for 5 min in this preparation and subsequent exposure at time ero to 8,uM hydrogen peroxide (line 6). The dotted line indicates the level at which the surviving fraction was.1, i.e., 9% of the cells were killed

3 VOL. 44, 1984 TOXICITY OF HYPOTHIOCYANITE-HYDROGEN PEROXIDE 583 hydrogen peroxide (Fig. 2). At lower concentrations of KSCN, the initial killing rates were higher than in the absence of KSCN. At 2 or 4,uM KSCN, however, the killing stopped after a few minutes, and a significant fraction of the cells survived (Fig. 2). When the cells of P. anaerobius were stored in a lactoperoxidase-free solution containing various concentrations of OSCN-, as little as 8,uM OSCN- potentiated the bactericidal effect of 8 p.m hydrogen peroxide (Table 1). In the presence of 32,uM lactoperoxidase-free OSCN-, 1,uM hydrogen peroxide had a bactericidal effect similar to that of 8,uM hydrogen peroxide alone (Table 1). When the lactoperoxidase-free OSCN- solution was mixed with hydrogen peroxide and the mixture was stored for various times before cells of P. anaerobius were added to it, the bactericidal effect of the mixture remained high, even after a 6-min storage, although the amount of OSCN- detected in the reaction mixture at that time was very low (Table 2). Sulfate, sulfite, cyanate, cyanide, carbonate, and ammonia have been postulated as products of the chemical or the enymic oxidation of thiocyanate by hydrogen peroxide (12, 15, 27, 38). These substances were not toxic to P. anaerobius at a concentration of.1 mm, and they did not potentiate the bactericidal effect of 8,uM hydrogen peroxide (data not shown). Cyanosulfurous and cyanosulfuric acids have also been postulated as products of these oxidations (6, 15, 3). Since there are no reliable assays for these substances, it was not possible to evaluate their possible role in the killing of the cells. The bactericidal effect of a mixture of OSCN--hydrogen peroxide was abolished by catalase and by lactoperoxidase 1..1 U- LL 5: 3 co.1 KSCN * VM -- * 4VM OPM 1OpM 5pM 2OPM I I I I I 2 4 6( TIME (min) FIG. 2. Killing of P. anaerobius after exposure to 8,M hydrogen peroxide in the presence of lactoperoxidase (1,ug ml-') and various concentrations of KSCN. TABLE 1. Killing of P. anaerobius by hydrogen peroxide in the presence of a lactoperoxidase-free OSCN- preparationa OSCN- Hydrogen Killing time (p.m) peroxide (min) (p.m) ± ± ± ± t ± ± ± 2. a Cells were stored for 5 min in.9 ml of MOPS buffer solution containing various concentrations of OSCN- before various amounts of hydrogen peroxide (.1 ml) were added. The killing time was defined as that time after the hydrogen peroxide addition at which the surviving fraction of cells was.1. Means ± standard deviations are given for three independent experiments. The killing time was determined in experiments similar to those described in the legend to Fig. 1. The killing time was defined by the intersection of the dotted line in Fig. 1 with the killing curve of the organism. (data not shown) but not by superoxide dismutase, methionine, tryptophan, histidine, or mannitol (Table 3). Reduced glutathione and 2-mercaptoethanol did not influence the bactericidal effect of 8,uM hydrogen peroxide, but they eliminated the potentiating effect of OSCN- on hydrogen peroxide toxicity (Table 3). DABCO significantly decreased the toxicity of the OSCN--hydrogen peroxide mixture (Table 3). This effect of DABCO would suggest that singlet oxygen mediated the bactericidal effect of the mixture. No oxygen could, however, be detected with an oxygen electrode when OSCN- in concentrations up to 4,uM was mixed with 1 mm hydrogen peroxide. The concentration of OSCN- decreased 7 ± 2% during a 1-h storage in MOPS buffer solution. DABCO did not influence the rate of OSCN- breakdown or the stability of hydrogen peroxide in MOPS buffer solution (data not shown). The bactericidal effects of OSCN--hydrogen peroxide mixtures were similar in MOPS buffer solution and phosphate-buffered saline (data not shown). E. coli was much more resistant to hydrogen peroxide than was P. anaerobius, but OSCN- also potentiated the bactericidal effect of hydrogen peroxide to this organism (Fig. 3). When HeLa cells were exposed to 3,uM hydrogen peroxide or to a mixture of 3,uM hydrogen peroxide and 2,uM of lactoperoxidase-free OSCN- in phosphate-buffered saline for 1 h, their capacity to proliferate was reduced significantly more by the OSCN--hydrogen peroxide mixture than by hydrogen peroxide alone (Fig. 4). DISCUSSION Lactoperoxidase in the presence of thiocyanate could protect bacteria (2, 8) and cultured mammalian cells (13) from killing by hydrogen peroxide. There are, however, also several reports in which lactoperoxidase in the presence of thiocyanate has potentiated the bactericidal effect of hydrogen peroxide (6, 7, 21, 23, 33, 36). The effect of lactoperoxidase-thiocyanate-hydrogen peroxide mixtures on bacteria is dependent on the experimental conditions. If the bacteria are cultured after the exposure to lactoperoxidase-thiocyanate-hydrogen peroxide on nutrient agar under aerobic conditions, they may not grow, whereas they grow readily on blood agar under anaerobic conditions

4 584 CARLSSON, EDLUND, AND HANSTROM TABLE 2. Killing of P. anaerobius by mixtures of various ages of hydrogen peroxide and a lactoperoxidase-free OSCNpreparationa Age of reaction OSCN- (,um) Killing time mixture (min) OSCin)M 12 < ± 5. 6 <1. 2. ± ± ± ± ± ± ± ± ± ± ±.4 a Hydrogen peroxide (.4 ml; 2,uM) was mixed with.4 ml of 35,uM OSCN- in MOPS buffer solution; at various times thereafter,.2 ml of a bacterial suspension was added. The levels of OSCN- in the reaction mixtures at the time of the bacterial addition were assayed in parallel reaction tubes. The definition of killing time is given in footnote a of Table 1. Means ± standard deviations are given for three independent experiments. The killing time was 39 ± 2.5 min when the bacteria were exposed to 8 F.M hydrogen peroxide in the absence of OSCN-. (2). The explanation may be that the OSCN- formed in the mixture oxidies bacterial sulfhydral groups (36), and the bacteria may not be able to rereduce these vital groups when cultured on nutrient agar under aerobic conditions. The present study demonstrated another situation, in which lactoperoxidase-thiocyanate-hydrogen peroxide was more bactericidal than hydrogen peroxide alone. This happened when hydrogen peroxide was in excess of thiocyanate in the mixture. There are several potentially toxic products that may be formed. In studies of the chemical oxidation of thiocyanate by hydrogen peroxide, Wilson and Harris (38) found sulfate, cyanate, carbonate, and ammonia as final products and postulated the following mechanisms: SCN- + H22-+ HOSCN + OH- (1) HOSCN + H22-* HOOSCN + H2 (2) HOOSCN + H22 -* H2SO3 + HOCN (3) HOCN + 2H2-* HC3- + NH4+ (4) H2SO3 + H22-- H2S4 + H2 (5) where reaction 1 is the rate-determining step. It is this reaction which is catalyed by lactoperoxidase (4, 19). From the stoichiometry of reactions in lactoperoxidase-thiocyanate-hydrogen peroxide mixtures it has been postulated that reaction 2 could be catalyed by lactoperoxidase and that cyanosulfurous acid (HOOSCN) could be further oxidied by hydrogen peroxide into cyanosulfuric acid (HO3SCN) (15, 29). HOOSCN + H22 -* HO3SCN + H2 (6) These compounds have been suggested to be more toxic than OSCN- (6). In the experiments in which the enymatic formation of these compounds has been postulated, the possibility was overlooked, however, that lactoperoxidase might have catalatic activity (9). Oxygen and OSCN- may be the only products of the lactoperoxidase-catalyed reaction between INFECT. IMMUN. hydrogen peroxide and thiocyanate at neutral ph and at a concentration of hydrogen peroxide which does not exceed that of thiocyanate. A highly toxic agent was, however, formed in the present study, in which the concentration of hydrogen peroxide exceeded that of thiocyanate in lactoperoxidase-thiocyanate-hydrogen peroxide mixtures. This agent was also formed in a reaction between OSCN- and hydrogen peroxide in the absence of lactoperoxidase. It was not possible to determine the nature of this agent. The postulated products of the chemical oxidation of OSCN- by hydrogen peroxide, sulfate, sulfite, cyanate, carbonate, and ammonia (38) were not bactericidal, and they did not potentiate the effect of hydrogen peroxide. The only other postulated product of the chemical oxidation of OSCN- by hydrogen peroxide is HOOSCN, and it stands out as the suspected killing agent. There was, however, no method available for demonstrating the presence of this compound in the reaction mixtures. In a recent polarographic study it has been demonstrated, however, that in addition to OSCN- another oxidation product of thiocyanate is formed in lactoperoxidasethiocyanate-hydrogen peroxide mixtures when the concentration of hydrogen peroxide exceeds that of thiocyanate (31). This product might be HOOSCN. The use of putative antagonists of various other possible toxic agents showed that hydroxyl radicals (mannitol), superoxide radicals (superoxide dismutase), or thiocyanogen anion radicals (tryptophan) (1) were not likely the effective molecular species. DABCO, a singlet oxygen quencher (32), decreased the toxicity of the OSCN--hydrogen peroxide mixture. The effective molecular species of the mixture could not possibly be singlet oxygen, since no oxygen was formed in these mixtures and the singlet oxygen traps, tryptophan, histidine, and methionine, did not have any protective effects. DABCO is, however, not specific for singlet oxygen. It also reacts with peroxy radicals (28). The protective effect of DABCO could not be ascribed to a reaction of DABCO with OSCN- or hydrogen peroxide, since OSCN- and hydrogen peroxide were as stable in the presence of DABCO as in its absence. Catalase completely annihilated the killing effect of the OSCN--hydrogen peroxide mixtures, and this could be ascribed to the decomposi- TABLE 3. Killing of P. anaerobius by mixtures (with added substances) of hydrogen peroxide and a lactoperoxidase-free OSCN- preparation' OSCN-H22 Addition Killing time mixture (min) + None 5. ± mm mercaptoethanol 32.1 ± mm reduced glutathione 38.5 ± mm mannitol 7.2 ± mm histidine 2.7 ± mm methionine 1.1 ± mm tryptophan 5.5 ± mm DABCO 25.3 ± 3.5 _ 8 pum H ± p.m H mm 35.5 ± 3.5 mercaptoethanol a Hydrogen peroxide (.5 ml; 1.6 mm) was mixed with.75 ml of 44 mm OSCN- in MOPS buffer solution. Various substances (1 p.1) were added 5 min after hydrogen peroxide and OSCN- had been mixed; after another 5 min,.2 ml of a bacterial suspension was added. The definition of killing time is given in footnote a of Table 1. Means ± standard deviations are given for three independent experiments.

5 VOL. 44, 1984 TOXICITY OF HYPOTHIOCYANITE-HYDROGEN PEROXIDE 585 tion of hydrogen peroxide in the mixture. Reduced glutathione and 2-mercaptoethanol decreased the toxicity of the mixture as expected, since these compounds readily convert OSCN- into thiocyanate (18, 35). To settle the nature of the killing agent, it is important that the actual products of the chemical oxidation of OSCN- by hydrogen peroxide and those of the lactoperoxidase-catalyed reaction between thiocyanate and hydrogen peroxide be fully identified. Since the lactoperoxidase-catalyed reaction has an optimum ph of 4.5 and oxygen is only formed at neutral and alkaline phs (9), the reaction products at various phs should be studied. In the absence of thiocyanate, lactoperoxidase potentiated the bactericidal effect of hydrogen peroxide. It is possible that various components of the bacterial cell or of the buffer solution served as substrates in reactions catalyed by lactoperoxidase in the presence of hydrogen peroxide, resulting in bactericidal products (2). The present results show that the biological effect of lactoperoxidase-thiocyanate-hydrogen peroxide mixtures will be determined by the actual concentration of the individual components. In the oral cavity, where the salivary secretions contain more than 1 mm thiocyanate and the production of hydrogen peroxide will rarely produce a higher concentration than.1 mm OSCN- (35), lactoperoxidase CO, -J -J w LL LU CD i TIME (days) FIG. 4. Number of HeLa cells per cm2 attached to the surface of culture flasks after 1 h of exposure to phosphate-buffered saline (line 1); 3 pfm hydrogen peroxide (line 2); or 3,uM hydrogen peroxide- 2,uM lactoperoxidase-free OSCN- (line 3). Exposure of the cells to 2,uM lactoperoxidase-free OSCN- or to a mixture of 3,uM hydrogen peroxide, 1,uM KSCN, and lactoperoxidase (1,ug ml-', produced results almost identical to those shown in line 1. 3 LL. 5; TIME (min) FIG. 3. Killing of E. coli after exposure to hydrogen peroxide or OSCN--hydrogen peroxide mixtures under anaerobic conditions, showing the surviving fraction of cells under the following conditions: after storage for 1 h in phosphate-buffered saline (line 1); in the presence of 5 p.m hydrogen peroxide (line 2); in the presence of 28 p.m OSCN- (line 3); in the presence of a mixture of 5 p.m hydrogen peroxide and 28 p.m lactoperoxidase-free OSCN- in phosphate-buffered saline (line 4). Hydrogen peroxide and lactoperoxidase-free OSCN- were mixed 1 min before the cells were added. 4 6 and thiocyanate could be considered as an efficient protective system of the mucous membranes against hydrogen peroxide toxicity. Hydrogen peroxide is not only converted into the less toxic OSCN-; this compound also stops the bacterial production of hydrogen peroxide (11). It is, of course, possible that this protective system may fail in individual cases, in which the lactoperoxidase or thiocyanate level of saliva is low, or the capacity of the bacteria at specific sites to produce hydrogen peroxide is very high. Lactoperoxidase-thiocyanate-hydrogen peroxide is usually referred to as an antimicrobial system of saliva. It is certainly true that OSCN- formed by this system reversibly stops glycolysis and consequently the acid production in many oral bacteria, but we feel that this effect is only a part of a more important function of salivary lactoperoxidase and thiocyanate, vi., to protect the salivary glands and the oral mucous membranes against hydrogen peroxide toxicity. ACKNOWLEDGMENTS This study was supported by the Swedish Medical Research Council (project no. 4977). LITERATURE CITED 1. Adams, G. E., J. E. Aldrich, R. H. Bisby, R. B. Cundall, J. L. Redpath, and R. L. Willson Selective free radical reactions with proteins and enymes: reactions of inorganic radical

6 586 CARLSSON, EDLUND, AND HANSTROM anions with amino acids. Radiat. Res. 49: Adamson, M., and J. Carlsson Lactoperoxidase and thiocyanate protect bacteria from hydrogen peroxide. Infect. Immun. 35: Ananthaswamy, H. N., and A. Eisenstark Repair of hydrogen peroxide-induced single-strand breaks in Escherichia coli deoxyribonucleic acid. J. Bacteriol. 13: Aune, T. M., and E. L. Thomas Aumulation of hypothiocyanite ion during peroxidase-catalysed oxidation of thiocyanate ion. Eur. J. Biochem. 8: Bachmann, B. J Pedigrees of some mutant strains of Escherichia coli K-12. Bacteriol. Rev. 36: Bjoerck, L., and. Claesson Correlation between concentration of hypothiocyanate and antibacterial effect of the lactoperoxidase system against Escherichia coli. J. Dairy Sci. 63: Bjorck, L., C.-G. Rosen, V. Marshall, and B. Reiter Antibacterial activity of the lactoperoxidase system in milk against pseudomonads and other gram-negative bacteria. Appl. Microbiol. 3: Carlsson, J Bactericidal effect of hydrogen peroxide is prevented by the lactoperoxidase-thiocyanate system under anaerobic conditions. Infect. Immun. 29: Carlsson, J Catalatic activity of lactoperoxidase in the presence of SCN-. Biochem. Biophys. Res. Commun. 116: Carlsson, J., G. P. D. Granberg, G. K. Nyberg, and M.-B. K. Edlund Bactericidal effect of cysteine exposed to atmospheric oxygen. Appl. Environ. Microbiol. 37: Carlsson, J., Y. Iwami, and T. Yamada Hydrogen peroxide excretion by oral streptocoi and effect of lactoperoxidasethiocyanate-hydrogen peroxide. Infect. Immun. 4: Chung, J., and J. L. Wood Oxidation of thiocyanate to cyanide and sulfate by the lactoperoxidase-hydrogen peroxide system. Arch. Biochem. Biophys. 141: Hainstrom, L., A. Johansson, and J. Carlsson Lactoperoxidase and thiocyanate protect cultured mammalian cells against hydrogen peroxide toxicity. Med. Biol. 61: Hoffman, M. E., and R. Meneghini Action of hydrogen peroxide on human fibroblast in culture. Photochem. Photobiol. 3: Hogg, D. McC., and G. R. Jago The antibacterial action of lactoperoxidase. Biochem. J. 117: Holdeman, L. V., P. Cato, and W. E. C. Moore (ed.) Anaerobic laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg. 17. Hoogendoorn, H The effect of lactoperoxidase-thiocyanate-hydrogen peroxide on the metabolism of cariogenic microorganisms in vitro and in the oral cavity. Mouton, Den Haag. 18. Hoogendoorn, H The inhibitory action of the lactoperoxidase system on Streptocous mutans and other microorganisms, p In H. M. Stiles, W. J. Loesche, and T. C. O'Brien (ed.), Microbial aspects of dental caries. Information Retrieval Inc., Washington, D.C. 19. Hoogendoorn, H., J. P. Piessens, W. Scholtes, and L. A. Stoddard Hypothiocyanite ion; the inhibitor formed by the system lactoperoxidase-thiocyanate-hydrogen peroxide. I. Identification of the inhibiting compound. Caries Res. 11: Kraus, F. W., J. F. Nickerson, W. I. Perry, and A. P. Walker Peroxide and peroxidogenic bacteria in human saliva. J. Bacteriol. 73: Law, B. A., and P. John Effect of the lactoperoxidase INFECT. IMMUN. bactericidal system on the formation of the electrochemical proton gradient in E. coli. FEMS Microbiol. Lett. 1: Mandel, I. D., J. Behrman, R. Levy, and D. Weinstein The salivary lactoperoxidase system in caries-resistant and -susceptible adults. J. Dent. Res. 62: Marshall, V. M. E., and B. Reiter Comparison of the antibacterial activity of the hypothiocyanite anion towards Streptocous lactis and Escherichia coli. J. Gen. Microbiol. 12: Mickelson, M. N Effect of lactoperoxidase and thiocyanate on the growth of Streptocous pyogenes and Streptocous agalactiae in a chemically defined culture medium. J. Gen. Microbiol. 43: Morrison, M., and W. F. Steele Lactoperoxidase, the peroxidase in the salivary gland, p In P. Persson (ed.), Biology of the mouth. American Association for the Advancement of Science, Washington, D.C. 26. Neidhardt, F. C., P. L. Bloch, and D. F. Smith Culture medium for enterobacteria. J. Bacteriol. 119: Oram, J. D., and B. Reiter The inhibition of streptocoi by lactoperoxidase, thiocyanate and hydrogen peroxide. Biochem. J. 1: Packer, J. E., J. S. Mahood, V.. Mora-Arellano, T. F. Slater, R. L. Willson, and B. S. Wolfenden Free radicals and singlet oxygen scavengers: reaction of a peroxy-radical with 1-carotene, diphenyl furan and 1,4-diaobicyclo(2,2,2)-octane. Biochem. Biophys. Res. Commun. 98: Pruitt, K. M., B. Mansson-Rahemtulla, and J. Tenovuo Detection of hypothiocyanite (OSCN-) ion in human parotid saliva and the effect of ph on OSCN- generation in the salivary peroxidase antimicrobial system. Arch. Oral Biol. 28: Pruitt, K. M., and J. Tenovuo Kinetics of hypothiocyanite production during peroxidase-catalyed oxidation of thiocyanate. Biochim. Biophys. Acta 74: Pruitt, K. M., J. Tenovuo, R. W. Andrews, and T. McKane Lactoperoxidase-catalyed oxidation of thiocyanate: polarographic study of the oxidation products. Biochemistry 21: Quannes, C., and T. Wilson Quenching of singlet oxygen by tertiary aliphatic amines. Effect of DABCO. J. Am. Chem. Soc. 9: Reiter, B., V. M. E. Marshall, L. Bjorck, and C.-G. Rosen Nonspecific bactericidal activity of the lactoperoxidase-thiocyanate-hydrogen peroxide system of milk against Escherichia coli and some gram-negative pathogens. Infect. Immun. 13: Riddles, P. W., R. L. Blakeley, and B. Zerner Ellmans's reagent: 5,5'-dithiobis(2-nitrobenoic acid)-a reexamination. Anal. Biochem. 94: Tenovuo, J., B. Mansson-Rahemtulla, K. M. Pruitt, and R. Arnold Inhibition of dental plaque acid production by the salivary lactoperoxidase antimicrobial system. Infect. Immun. 34: Thomas, E. L., and T. M. Aune Lactoperoxidase, peroxide, thiocyanate antimicrobial system: correlation of sulfhydryl oxidation with antimicrobial action. Infect. Immun. 2: White, W. E., Jr., K. M. Pruitt, and B. Mansson-Rahemtulla Peroxidase-thiocyanate-peroxide antibacterial system does not damage DNA. Antimicrob. Agents Chemother. 23: Wilson, I. R., and G. M. Harris The oxidation of thiocyanate ion by hydrogen peroxide. II. The acid catalyed reaction. J. Am. Chem. Soc. 83: