Quick viewing(Text Mode)

Biological Reactivity of Hypochlorous Acid

Biological Reactivity of Hypochlorous Acid

Proc. Nati Acad. Sci. USA Vol. 78, No. 1, pp. 210-214, January 1981 Biochemistry

Biological reactivity ofhypochlorous : Implications for microbicidal mechanisms ofleukocyte (bactericide/biological oxidation//respiration/toxin) J. MICHAEL ALBRICH, CAROL A. MCCARTHY, AND JAMES K. HURST* Department ofChemistry and Biochemical Sciences, Oregon Graduate Center, Portland, Oregon 97006 Communicated by Henry Taube, October 6, 1980

ABSTRACT Oxidative degradation ofbiological substrates by electrophile-nucleophile interactions involving association of hypochlorous acid has been examined under reaction conditions electropositive with electron-rich centers on the sub- similar to those found in active phagosomes. Iron sulfur are bleached extremely rapidly, followed in decreasing order by strate; reaction pathways are-correspondingly highly dependent f3carotene, nucleotides, porphyrins, and heme proteins. En- upon medium-conditions, particularly H' and Cl- concentra- zymes containing essential molecules are inactivated with tions (14). With the exception ofamines and amino (11, 12, an effectiveness that roughly parallels the nucleophilic reactivities 15), the reactions of biological compounds with HOC1 are not of their sulfhydryl groups. Other compounds, including glucos- understood (16). , qumiones, riboflavin, and, except for N-chlorination, We report 'here the results of a survey of simple biological phospholipids, are unreactive. Rapid irreversible oxidation ofcy- compounds which can be taken as prototypic ofvarious tochromes, adenine nucleotides, and carotene pigments occurs compo- when bacterial cells are exposed to exogenous hypochlorous acid; nents of bacterial cells.:The data demonstrate that HOCI is with , titrimetric oxidation of was strongly selective -toward nucleotides and compounds that are found to coincide with loss ofaerobic respiration. The occurrence models for certain components ofrespiratory redox chains. This of these cellular reactions implicates hypochlorous acid as a pri- selectivity is shownito extend to bacterial cells. mary microbicide in myelo.roxidase-containing leukocytes; the reactivity patterns observed are consistent with the view that bac- MATERIALS AND METHODS tericidal action results primarily from loss ofenergy-linked respi- ration due to destruction ofcellular electron transport chains and Reagents. HOCI was prepared by vacuum distilling Chlorox the adenine nucleotide pool. (NaOCl) after adjusting the pH to 7-8 with dilute ; concentrations were determined by spectrophotometric [E2W commonly lose their ability to divide within minutes of = 100 (17)] or iodometric'analyses. Ethanolchloramine was pre- encountering phagocytosing leukocytes (1). Loss ofcell viability pared by reaction ofHOC1 with a 3-fold excess ofethanolamine often occurs well before the onset ofcellular digestion as deter- in 0.1 M phosphate buffer at pH 5.0. Aldolase (EC 4.1.2.13) was mined by physiological changes (2), macromolecular degrada- prepared from brewer's yeast (Standard Brands) by ammonium tion (3, 4), or loss of macromolecular biosynthesis (5). The spe- sulfate precipitation (18) in the presence of proteolysis inhibi- cific reactions giving rise to cellular death have not yet been tors (19); the purified had a specific activity of 64 units/ identified, at least in part because leukocytes are capable ofini- mg of . Other and reagent-grade biological tiating a diverse set of processes which are potentially lethal (6, compounds were used as obtained from commercial suppliers. 7). Given the above observations, however, the microbicidal re- Escherichia coli (ATCC 25922) were grown on media 2 (20) plus actions must be among the first that attend interaction with the succinate (2 g/liter) at pH 6.5 and harvested in stationary phase. leukocyte. The bacteria were centrifuged, washed, and resuspended in Reactions catalyzed by myeloperoxidase (MPOase) appear buffer immediately before use. Sarcina lutea (Mwrococcus lu- to make major contributions to the microbicidal-action of poly- teus, generously provided by S. J. Klebanoff, University of morphonuclear leukocytes (PMNs) (6, 7). The cell-free Washington Medical School, Seattle) were grown on 5% tryptic MPOase-H202-Cl- system is 'potentially microbicidal; chlori- soy agar plates (Prepared Media Laboratories); suspensions nation of bacteria by the cell-free system (8, 9) and of macro- were made by carefully scraping the cells from the plates and molecular fractions during PMN digestion of bacteria (9) sug- agitating them in buffer. Bacterial cell concentrations were de- gests that the toxic mechanisms may involve,direct reaction with termined by phase-contrast microscopy with a hemocytometer. hypochlorous acid (HOCI), the oxidized product of MPOase HOCI Oxidations. Reaction with biological compounds was catalysis. Because HOCI is freely diffusible from the enzyme monitored byusingeither a Cary 16 spectrophotometer or Gib- (10) and the gross features of cellular reactions with phagocy- son-type stopped-flow apparatus (21). Rate laws were deter- tosing PMNs, the MPOase-H202Cl system, and exogenous mined from initial rates ofbleaching; rate constants for the com- HOC1 are similar (4, 8, 11-13), it is likely that the actual mi- pounds listed in Table 1 were -determined from loss of their crobicidal reactions are uncatalyzed. If so, minimal require- characteristic absorption bands when HOC1 was in large excess. ments for toxicity would seem to be that HOCl show high se- The reactions are generally complex, involving numerous oxi- lectivity in its reactions with biological substrates and that its dized intermediates. Any rapid initial degradative steps that did action be directed against susceptible sites-e.g., specific en- not give rise to large spectral changes (e.g., heterocyclic ring ergy-generating or biosynthetic systems. chlorination -would go undetected. Consequently, the rate con- In general, rates 'of oxidation and chlorination by various stants reported would be lower limits for these steps. Enzyme monovalent chlorine compounds can be understood in terms of inactivation was studied by addingportions ofHOC1 to solutions

The publication costs ofthis article were defrayed in part by page charge Abbreviations: MPOase, myeloperoxidase; PMN; polymorphonuclear payment. This article must therefore be hereby marked "advertise- leukocyte. ment" in accordance with 18 U. S. C. §1734 solely to-indicate this fact. * To whom reprint requests should be addressed. 210 Downloaded by guest on September 28, 2021 Biochemistry: Albrich et aL Proc. NatLAcad. Sci. USA 78 (1981) 211 containing freshly dialyzed enzyme and then measuring, by ap- =12:1. Immediate reaction products had masses that increased propriate standard assay procedures, the residual activity ofali- in units of 32 ± 1 atomic mass units, suggesting epoxidation quots taken from the reaction solution. In the enzyme and cel- through intermediary formation of chlorohydrins. Because re- lular oxidation studies, HOCI was added as a bolus to rapidly tinol oxidation by HOCl gives products with masses increasing stirred reactant solutions. HOCI-promoted loss ofbacterial res- by 16 ± 1 atomic mass units, the opposite ends of the carotene piration was measured with a Clark electrode and recorder as- molecule may be reacting largely independently. /-Carotene is sembly (14). also bleached by the MPO-H202-Cl- system (24). Optical difference spectra of bacterial suspensions were re- Oxidation of numerous porphyrins, hemes, and hemopro- corded on an Aminco DW-2 spectrophotometer. Cellular nu- teins, including protoporphyrin IX, mesoporphyrin IX di- cleotide concentrations were determined, by using an Aminco methyl ester, 2-formyl4-vinyl-deuteroporphyrin dimethyl es- Chem-Glow photometer, from the integrated light intensity of ter, 2,4-diformyldeuteroporphyrin dimethyl ester, the luciferin-luciferase assay (22) taken over the first 3 sec ofre- ferriprotoporphyrin IX (hemin) and its dimethyl ester, ferridi- action. In the nucleotide studies, reactions after exposure to acetyldeuteroporphyrin, (III) mesoporphyrin IX di- HOCl were quenched by rapidly withdrawing samples into a methyl ester, myoglobin, microperoxidase, and cytochrome c, spring-loaded syringe containing . Intervals be- also occurred rapidly. Porphyrin and heme solutions contained tween mixing and quenching (10-300 sec) were chosen so that 1% NaDodSO4 to solubilize the macrocycle. Cytochrome c oxi- reaction with HOCI was completed and subsequent intracellu- dation by the cell-free MPO-H202-Cl- system is also rapid (un- lar reactions were minimized. In several experiments, the re- published data). action was quenched with sodium thiosulfate before acidifica- Of the porphyrin compounds examined in detail (protopor- tion to ensure that oxidation was not due to molecular chlorine phyrin IX, mesoporphyrin IX dimethyl ester, hemin, and cyto- formed from residual HOCL. The results were identical to those chrome c), all except protoporphyrin IX exhibited bleaching for runs in which thiosulfate quenching was omitted. Mass spec- rates that increased with increasing H' concentration in the pH tra were recorded on a Hitachi-Perkin-Elmer RMU-7 instru- region 4-7, although the dependency was complex and less than ment operated in the field desorption mode. first-order; protoporphyrin IX rates were inversely dependent upon acidity in this region. Reactions ofhemin and cytochrome RESULTS c were first order in both reactants. Oxidation ofmetal-centered Oxidation of Biological Substrates. Degradative oxidation of porphyrin compounds was Cl- independent; that ofmetal-free porphyrins was linearly dependent upon Cl- concentration. various biological compounds that contain electron-rich func- Bleaching stoichiometries increased from HOC/porphyrin tional substituents was rapid. In each instance the susceptible 4, 10, 12, and 24 for the compounds myoglobin, hemin, micro- site comprised a ir-conjugated or similar electron delocalized The center. Examples from each reaction type are given in Table 1. peroxidase, and cytochrome c, respectively. progressive Ferredoxin oxidation occurred at rates too fast to measure by increase in stoichiometry through the series was associated using stopped-flow methods, even at submillimolar reactant roughly with decreasing accessibility of the porphyrin ring to concentrations. The rate constants listed for these reactions solvent. Addition of excess HOCl to solutions containing cy- were calculated by assuming that the rate law is first-order in tochrome c gave sequential bleaching and precipitation, indi- each reactant; no Cl- dependence was observed over the range cating that reaction with HOCl is selective for the porphyrin investigated (1-100 mM, pH 5). With spinach ferredoxin, oxi- ring, at least to the extent that reaction at other sites is insuf- dation occurred with complete loss of the iron-sulfur chromo- ficient to cause extensive protein denaturation. Reaction of phore at stoichiometric ratios ofabout 10 HOCl/Fe2S2 unit; the hemes and hemoproteins with peroxide and ethan- product spectrum was identical to that of apoprotein (23), sug- olchloramine was found to occur at rates no more than '/Aoooth gesting oxidation of sulfur with probable loss of iron. Chromo- those with HOCl. phoric bleaching of ferredoxins initiated by MPOase-catalyzed Oxidation ofadenine nucleotides was first order in substrate, peroxidation ofCl- is also rapid (unpublished data). HOCl, and Cl-; the H+ dependence was complex, with rates Oxidation of /3-carotene is first-order in both reactants and increasing with acidity and exhibiting less than first-order be- first-order in Cl- concentration; rates increased only slightly havior. The initial reactions appeared to involve ring chlorina- with H+ concentration in the weakly acidic region (pH 4-7). tion, as suggested by the bathochromic shift of the UV absor- Complete bleaching occurred at a ratio of HOCl to carotene of bance maximum (25). Subsequent loss ofthe UV chromophore, which was rapid (Table 1), indicates oxidative destruction ofthe heterocyclic ring. Complete bleaching required addition of Table 1. Rate constants for oxidative bleaching of biological HOCl to a ratio ofHOCl/AMP 10. substrates by HOCI Introduction of electron-withdrawing substituents onto aro- Compound A, nm k, (M-'s1-) x 10-3 matic or heterocyclic rings was found to decrease their reactivity IDP at rates Spinach ferredoxin 400 ¢100* toward HOCL. GMP, CDP, and reacted comparable Clostridial ferredoxin 420 40 to the rate for ADP; UDP was less rapid. Riboflavin and a qui- (-Carotene in 1% NaDodSO4t 459 23 none (menadione, ubiquinone-10 in 1% NaDodSO4) were to- Cytochrome c 409 2 tally unreactive under the experimental conditions (Table 1). Hemin-Ci in 1% NaDodSO4t 394 2 Hydroquinone was oxidized only to the level of benzoquinone; 280 2 no chlorine substitution or other evidence of oxidative degra- Amines and amino acids - 0.1-4t dation was found by spectrophotometric or mass spectrometric - § analyses. Oxidation rates were rapid; measured rate constants were comparable to those for cytochrome and nucleotide Reactions were in 0.05 M phosphate at pH 4.5; Cl- = 0.10 M; tem- bleaching. perature = 230C. * Too fast to measure. Compounds that can be taken to represent cell wall compo- t NaDodSO4 is unreactive. nents (N-acetylglucosamine, N-acetylmuramic acid, , * Calculated from data in ref. 15. and mannose) and, with the exception of ethanolamine which t'12= 7 sec (14). forms ethanolchloramine, those representing structural cell Downloaded by guest on September 28, 2021 212 Biochemistry: Albrich et al. Proc. Nad Acad. Sci. USA 78 (1981) membrane components (phosphatidycholine, choline, and glyc- erol) were unreactive over periods of up to several hours after mixing with HOC. Enzyme Inactivation. Reaction ofHOC1 with cysteine appar- ently occurs with preferential side-chain oxidation; no chlora- mine formation was noted in studies in which sulfhydryl oxida- tion was carried to the level ofsulfonic acid (12). Loss ofcatalytic A Abs. function resulting from enzyme sulfhydryl oxidation has been T suggested to be the microbicidal event, both from studies on en- 0.02 zyme inactivation (26) and from studies correlating loss of cell viability with apparent loss of titrable sulfhydryl groups in E. coli (8, 13). To examine these putative relationships more closely, we measured HOCI inactivation ofseveral enzymes with different sulfhydryl functions. Of the enzymes studied, papain (EC 3.4.4.10), glyceraldehyde-3-phosphate dehydrogenase (EC 400 440 480 520 560 600 640 1.2.1.12), and alcohol dehydrogenase (EC 1.1.1.1) have excep- A, nm tionally reactive active-site cysteine groups (27-29), yeast aldol- ase (EC 4.1.2.6) and lactate dehydrogenase (EC 1. 1. 1.27) have FIG. 1. HOCI oxidation ofE. coli. Conditions: 5 x 109 cells per ml suspended in 0.1 M phosphate buffer at pH 5.0; Cl- = 0.1 M; tempera- essential which are not the most reactive in the protein ture = 23°C. Curve a: difference spectrum, resting cells vs. respiring (30, 31), and muscle aldolase contains catalytically active cys- cells; curves b-e, incremental additions of50 ,ul of60 mM HOCl to the teine groups classed as having only auxiliary function (32). The reference cell, with an equal volume ofbuffer added to the sample cell. enzymes and HOC1 were allowed to react in 0.05-0.10 M phos- Respiration ceased with addition of150 1,u ofHOCI (curve d). The titra- phate buffer containing 0.10 M Cl- at pH 7 and in weakly acidic tion was carried to 500 ,ul total added HOC1 (not shown) without fur- media. The pH range was limited by irreversible loss ofactivity ther change in difference spectrum intensities. which occurred at more acidic conditions. In general, pH values of4-5 were accessible, although for alcohol dehydrogenase the 429, 529, and 558 nm with HOC1 addition are indicative ofoxi- limiting pH was 6.1. Over the ranges investigated, inactivation dation of b-type (34-36). Reaction was complete kinetics appeared to be independent of pH. In general, loss of within the time interval between mixing and recording ofspec- activity was not linearly proportional to the amount of added tra (1-2 min). Amplitudes ofthe difference peaks increased pro- HOCl; HOC1 became less effective with increasing extent ofthe portionately with HOC1 addition to the point of complete inhi- reaction. For this reason, stoichiometries to 50% total inactiva- bition of respiration. Continued addition of HOCl caused no tion have been reported (Table 2). Sensitivity to HOC1 inacti- further change in the visible region. vation followed nucleophilic reactivity of catalytically active It was not possible to distinguish between simple oxidation to sulfhydryl groups, providing some circumstantial evidence for the ferric state and oxidative destruction ofthe hemoporphyrin their involvement. It is clear from the data, however, that HOC1 from the spectral data because the cytochrome spectrum is ob- oxidation of side-chain groups is fairly nonselective. The rela- scured by larger optical changes occurring below 420 nm, pre- tively high HOCl/enzyme inactivation ratios, nonlinear titra- sumably attributable to changes in light-scattering properties of tion behavior, and absence ofany discrimination between reac- the cells accompanying oxidation. However, dithionite was in- tivities in essential and nonessential cysteine groups in the capable ofreversing the effects ofreaction with HOCl, suggest- titration curves-e.g., for yeast aldolase (30)-all support this ing that cytochrome bleaching had occurred. In contrast, oxi- contention. dation by ethanolchloramine or H202 gave a nearly identical Oxidation of Bacterial Cells. In accordance with others (13, 26, 33) we found that HOC1 caused rapid loss ofoxygen uptake by respiring bacteria. The spectraofE. coli to which incremental additions of HOC1 were made are compared with those of un- treated bacteria in Fig. 1. Under the experimental conditions the cytochromic components of resting cells exist almost com- pletely in their reduced oxidation states; addition of dithionite ~Q06 to either cuvette in the absence of HOC1 gave no change in the ¢1,.e baseline difference spectrum. The spectral peaks appearing at

Table 2. Enzyme inactivation by HOCI SH groups, no./ HOCI/catalytic unit catalytic unit to 50% inactivation, Enzyme* Essential Total mol/mol 640 A, nm Papain 1 1 1.0 ADH 2 8-9 1 FIG. 2. Difference spectrum titration of S. lutea with HOCl. Con- LDH 1 4 -2.5 ditions: 0.05 M phosphate at pH 4.45; Cl- = 0.1 M; temperature = GAPDH 1 3 3 23'C. Sequential additions of 30 ,ul of 17.5 mM HOCI to the reference Aldolase (class II) 2 4 "'6 cell containing 3.0-ml suspension of9 x 108 coccal pairs ofS. lutea per Aldolase (class I) 0 8 -23 ml; an equal volume of buffer was added to the sample compartment. Difference curves are identical to carotene absorption spectra. (Inset) * ADH, alcohol dehydrogenase; LDH, lactate dehydrogenase; GAPDH, Titrimetricbehavior at447 nm; optical changes havebeencorrected for glyceraldehyde-3-phosphate dehydrogenase. dilution by titrant. Downloaded by guest on September 28, 2021 Biochemistry: Albrich et aL Proc. Natd Acad. Sci. USA 78 (1981) 213

Table 3. Adenine nucleotide levels in HOCI-treatedE. coli Nucleotide Adenylate Reaction time, Integrated intensity (relative)* concentrations energy sec ATP ATP & ADP ATP & ADP & AMP (relative) charge 0 (control) 19 31 43 1.0 0.58 10 6 8 17 0.40 0.41 35 8 12 19 0.44 0.53 300 6 8 16 0.37 0.43 5 x 108 cells per ml; HOCl = 90 81M in 0.05 M phosphate buffer at pH 4.5; Cl- = 0.10 M; temperature = 230C. HOCI concentration wasjust sufficient to inhibit respiration. * Luciferin-luciferase assay as described in text (22). difference spectrum which was lost upon dithionite addition, These types ofreactions also appear to take place during oxi- indicating fully reversible oxidation with these reagents. dation ofbacterial cells. Respiratory loss-in E. coli that occurs as Exposure ofS. lutea to HOCI gave rise to oxidative bleaching a consequence of oxidation by HOCI (13, 26, 33) or the of its carotenoid components (Fig. 2); reaction was instanta- MPO-H202-CL- system (6) has been shown to coincide, more neous and led to total loss ofpigmentation. or less, with loss ofcell viability (13, 26, 33, 39). Irreversible ox- Adenine nucleotide oxidation also occurred within seconds of idation ofbacterial cytochromes also coincides with respiratory mixing E. coli with HOC. Approximately two-thirds ofthe total loss (Fig. 1). Although the exact sites of reaction remain to be nucleotide concentration was lost upon addition ofa quantity of determined, the correlations suggest a probable link between HOC1 just sufficient to inhibit respiration; the reaction was loss ofelectron transport function and cellular death. Our con- complete within 10 sec after mixing (Table 3). tention that bacterial cell walls do not impose any substantial barrier to attack by HOCI is supported by the direct observa- DISCUSSION tions ofreactant bleaching in the cell membrane (carotene, Fig. 2) and possibly the cytosolt (adenine nucleotides, Table 3). Loss Phagocytosis stimulates -linked respiration, generat- of ATP resulting from HOCI oxidation of E. coli has also been ing hydrogen peroxide in . Bacterial engulfment is reported by others (33), although itwas not determined whether accompanied by migration of leukocytic granules, containing phosphate ester or purine oxidation had occurred. among their constituents the enzyme MPOase, to the site ofthe Our studies show that, when sufficient HOCI is added to just forming phagosome. During subsequent fusion of granule and inhibit respiration, the predominant effect is oxidative degra- phagosomal cell walls, the granule proteins are injected into the dation, not loss of adenylate energy charge-e.g., that might phagosome. As degranulation proceeds, the phagosome be- occur as a consequence ofimpaired capacity for oxidative phos- comes progressively more acidic. The entire process from the phorylation. point of encounter of bacterium with PMN to completion re- The biological compounds we have studied are generally quires only several minutes and leaves the bacterium isolated markedly less reactive toward other types ofoxidants formed in within a vacuole containing MPOase and an enzymatic H202- phagocytosing leukocytes. In a cursory survey, we find no ap- generating system in a medium containing 0.1 M Cl- at pH 4-6 preciable reaction of hemes, carotene, or AMP with H202 or (6, 7); these conditions are optimal for MPOase-catalyzed gen- ethanolchloramine. Spinach ferredoxin is oxidized only slowly eration of HOC. As noted above, the time frame for phagocy- by H202 but is rapidly bleached by the chloramine. The possi- tosis is the same as that required for loss ofcell viability. None- bility exists, therefore, that HOCI oxidation of bacterial iron- theless, bacterial digestion within the leukocyte continues for sulfur centers might be mediated by.formation of endogenous several hours. chloramines (8, 13). Hemes are unreactive toward oxidative The reactivity patterns found here indicate that cytochromes, degradation by superoxide ion (41, 42); from consideration ofits iron-sulfur proteins, and nucleotides are highly vulnerable to chemical reactivity (42), superoxide ion can be anticipated to be oxidative degradation by HOC1. Because these types of com- a poor oxidant for the other compounds as well. pounds appear essential to virtually all aerobic and anaerobic To summarize briefly, the character of the biological oxida- electron transport systems (37, 38), their loss undoubtedly tions described here is appropriate for effective inactivation of would be lethal. The cell envelope appears to afford little pro- cells, and reactions occur on time scales associated with cellular tection against oxidation by exogenous HOC1 because, except death; other leukocytic oxidizing agents appear generally not to for amines, other structural cell wall and membrane com- meet these criteria. We therefore propose that HOCl is the mi- ponents are unreactive. It is noteworthy that oxidation ofmetal- crobicidal agent resultingfrom MPOase-catalyzed reactions and loporphyrins and ferredoxins is chloride-independent; reaction it exerts its toxic action by oxidizing or oxidatively destroying ofsimilar redox centers bound within the lipophilic membrane electron transport chains and the adenine nucleotide pool (ATP, interior is unlikely to be limited by the low levels ofCl- at those ADP, and AMP). The vulnerability of these systems to HOCl sites. Furthermore, the increase in phagosomal acidity that ac- suggests a universal mechanism of leukocyte toxicity toward companies PMN degranulation serves to enhance selectivity for fungi and mammalian cells [particularly tumor cells (43)] as well these compounds over amines. Nucleotide, heme, and hemo- as bacteria. protein bleaching rates increase with acidity, whereas chlora- Alternative proposals for killing mechanisms have invoked mine formation rates decrease, so that in neutral to weakly various consequences of sulfhydryl oxidation (13, 26, 44), cell acidic media the former are more rapid (Table 1). HOCI oxida- envelope alteration (4, 33, 39), or nucleotide dysfunction (45, tion of H202, which is relatively slow (14), is also minimized in 46). These hypotheses were made either on the basis of evi- weakly acidic solution. The reactive compounds listed in Table 1 are therefore representative members of the only biological t Reported turnover rates for microbial ATP pools are sufficiently large classes yet identified which undergo rapid degradative oxida- that several cycles can occur within the times of exposure of cells to tion by HOCI under conditions similar to those existing within HOCI (40). The possibility exists, therefore, that nucleotides are the PMN phagosome. bleached only when membrane-bound. Downloaded by guest on September 28, 2021 214 Biochemistry: Albrich et al. Proc. Natl. Acad. Sci. USA 78 (1981) dence demonstrating reaction in homogeneous solution or the 10. Harrison, J. E. & Schultz, J. (1976)J. Biol Chem. 251, 1371-1374. occurrence of reactions in cells after relatively long periods of 11. Stelmaszynska, T. & Zgliczynski, J. M. (1978) Eur. J. Biochem. 92, incubation with HOCI, the MPO-H202-CL- system, or PMN 301-308. 12. Pereira, W. E., Hoyano, Y., Summons, R. E., Bacon, V. A. & Duf- granules. There is little doubt that digestion under these con- field, A. M. (1973) Biochim. Biophys. Acta 313, 170-180. ditions will lead to sulfhydryl oxidation and N-chlorina- 13. Thomas, E. L. (1979) Infect. Immun. 25, 110-116. tion (8, 13), but the role of these reactions in bactericidal action 14. Held, A. M., Halko, D. J. & Hurst, J. K. (1978)J. Am. Chem. Soc. is unclear. For instance, enzyme sulfhydryl oxidation rates are 100, 5732-5740. unknown but are apparently insufficiently rapid to compete 15. Morris, J. C. (1967) in Principles and Applications of with nucleotide reactions in the . Camper and McFeters Chemistry, eds. Faust, S. D. & Hunter, J. V. (Wiley, New York), pp. 23-53. (33) have shown that HOC1 disinfection ofE. coli causes a drastic 16. Held, A. M. & Hurst, J. K. (1978) Biochem. Biophys. Res. Com- decrease in ATP levels but no perceptible loss of aldolase activ- mun. 81, 878-885. ity, despite our studies which show that class II aldolases are 17. Morris, J. C. (1966)J. Phys. Chem. 70, 3798-3805. among the enzymes most sensitive to HOC1 inactivation (Table 18. Rutter, W. J. & Hunsley, J. R. (1966) Methods Enzymol 9, 2). Although loss ofcell viability is reported to correlate with loss 480-486. of titrable sulfhydryl groups in E. coli (8, 13), application of the 19. Mildvan, A. S., Kobes, R. D. & Rutter, W. J. (1971) Biochemistry 10, 1191-1204. titrant used [5,5'-dithio-bis(2-nitrobenzoic acid)] to this analysis 20. Stanier, R. Y., Adelberg, E. A. & Ingraham, J. (1976) The Micro- is questionable. Both ferrous salts (52) and hydroquinones (un- bial World(Prentice-Hall, Englewood Cliffs, NJ), 4th Ed., p. 163. published data) are capable of reducing this titrant. Because one 21. Lane, R. H. & Hurst, J. K. (1974) Biochemistry 13, 3292-3297. likely consequence of HOCl-induced loss of respiration is ox- 22. Ching, T. M. & Ching, K. K. (1972) Plant Physiol 50, 536-540. idation of the electron transport chains (Fig. 1), it is unclear to 23. Lovenberg, W., Buchanan, B. B. & Rabinowitz, J. C. (1963)j. Biol what extent the differences in this reaction with HOCl-treated Chem. 238, 3899-3913. 24. Harrison, J. E., Watson, B. D. & Schultz, J. (1978) FEBS Lett. 92, and untreated E. coli are due to loss of reducing equivalents in 327-332. respiratory redox chains as opposed to bona fide sulfhydryl 25. Montgomery, J. A. & Holum, L. B. (1975)J. Am. Chem. Soc. 79, group oxidation. 2185-2188. Similarly, simple chloramines are effective microbicides (13), 26. Knox, W. E., Stumpf, P. K., Green, D. E. & Auerback, V. H. but those formed by N-chlorination of taurine and other biolog- (1948)j. Bacteriol. 55, 458-460. ical amines and amino acids are not toxic (47-49) unless incu- 27. Torchinskii, Y. M. (1974) Sulphydryl and Groups ofPro- bated with the microbes for several hours teins (Consultants Bureau, New York), pp. 126-129. (13). Because chlora- 28. Harris, J. I. & , M. (1976) The Enzymes, ed. Boyer, P. mines derived from endogenous amines show this same (Academic, New York), Vol. 13, pp. 1-49. behavior (8, 13), their role in physiological disinfection pro- 29. Branden, C. I., Jornvall, H., Eklund, H. & Furugren, B. (1975) cesses is probably minimal. One point which our studies do not The Enzymes, ed. Boyer, P. (Academic, New York), Vol. 11, pp. address is the loss of transport function that occurs when bacte- 103-190. rial cells are inactivated by HOC1 (33, 50); both HOC1 (8) and 30. Ingram, J. M. (1969) Can. J. Biochem. 47, 595-601. the system (4, 8) 31. Holbrook, J. J., Liljas, A., Steindel, S. J. & Rossman, M. G. (1975) MPO-H202-CL- have been shown to react The Enzymes, ed. Boyer, P. (Academic, New York), Vol. 11, pp. with cell envelope proteins. 191-292. Finally, singlet formed by HOC1 oxidation of H202 32. Steinman, H. M. & Richards, F. M. (1970) Biochemistry 9, has been discussed as a possible phagosomal toxin (6, 7). Its rate 4360-4372. of formation appears to be too slow to account for disinfection, 33. Camper, A. K. & McFeters, G. A. (1979)AppL Environ. Microbiol. however (Table 1) (16). Viewed from our perspective, the pro- 37, 633-641. tective effect of /3-carotene noted in studies with S. lutea mu- 34. Downie, J. A. & Cox, G. B. (1978) J. Bacteriol 133, 477-484. 35. Itakaki, E. & Hager, L. P. (1966)J. Biol. Chem. 241, 3687-3695. tants (51) is ascribable not to its ability to quench singlet oxygen, 36. Kita, K., Yamato, I. & Anraku, Y. (1978) J. Biol Chem. 253, as has been proposed, but to its ability to rapidly scavenge HOC1 8910-8915. in the cell membrane, thereby preventing loss of essential cel- 37. Haddock, B. A. & Jones, C. W. (1977) Bacteriol Rev. 41, 47-99. lular function. 38. Yoch, D. C. & Carithers, R. P. (1979) Microbiol Rev. 43,384-421. 39. Venkobachar, C., lyengar, L. & Rao, A. V. S. P. (1977) Water Res. We express our gratitude to Dr. Norman Bishop and Jim Wong (Or- 11, 727-729. egon State University) for making available to us their Aminco DW-2 40. Knowles, C. J. (1977) in Microbial Energetics, eds. Haddock, B. instrumentation facility, to Dr. Te May Ching (Oregon State University) A. & Hamilton, W. A. (Cambridge Univ. Press, London), pp. for use ofinstruments and for her generous assistance in making the nu- 241-284. cleotide measurements, and to Dr. Peter Hambright (Howard Univer- 41. Pasternack, R. F. & Halliwell, B. (1979)J. Am. Chem. Soc. 101, sity) for his gift of various porphyrins and metalloporphyrins. We are 1026-1031. deeply indebted to Drs. Seymour Klebanoff and Henry Rosen (Univer- 42. Valentine, J. S. (1979) in Biochemical and Clinical Aspects of Ox- sity ofWashington) for many helpful discussions. This research was sup- ygen, eds. Caughey, W. S. & Caughey, H. (Academic, New York), ported by grants from the Medical Research Foundation of Oregon and pp. 659-677. the National Institutes of Health (AI-15834). 43. Slivka, A., LoBuglio, A. F. & Weiss, S. J. (1980) Blood 55, 347-350. 1. Elsbach, P. (1973) N. EngLJ. Med. 16, 846-852. 44. Venkobachar, C., Iyengar, L. & Rao, A. V. S. P. (1975) Water Res. 2. Ayoub, E. M. & White, J. G. (1969)J. Bacteriol. 98, 728-736. 9, 119-124. 3. Cohn, Z. A. (1968)J. Exp. Med. 117, 27-42. 45. Ingols, R. S. (1958) Public Health Works 89, 105-106. 4. Selvaraj, R. J., Paul, B. B., Straus, R. R., Jacobs, A. A. & Sbarra, 46. Shih, K. L. & Lederberg, J. (1976)J. Bacteriol 125, 934-945. A. J. (1974) Infect. Immun. 9,255-260. 47. Paul, B. B., Jacobs, A. A., Strauss, R. R. & Sbarra, A. J. (1970) 5. Elsbach, P., Pettis, P., Beckerdite, S. & Franson, R. (1973)J. Bac- Infect. Immun. 2, 414-418. teriol 115, 490-497. 48. Strauss, R. R., Paul, B. B., Jacobs, A. A. & Sbarra, A. J. (1971) 6. Klebanoff, S. J. & Clark, R. A. (1978) The Neutrophil: Function Infect. Immun. 3, 595-602. and Clinical Disorders (Elsevier, North-Holland, Amsterdam). 49. Jacobs, A. A., Paul, B. B., Strauss, R. R. & Sbarra, A. J. (1970) 7. Babior, B. M. (1978) N. EngLJ. Med. 298, 659-666; 721-725. Biochem. Biophys. Res. Commun. 39, 284-289. 8. Thomas, E. L. (1979) Infect. Immun. 23, 522-531. 50. Klebanoff, S. J. (1968)J. Bacteriol 95, 2131-2138. 9. Zgliczynski, J. M. & Stelmaszynska, T. (1975) Eur.J. Biochem. 56, 51. Krinsky, N. I. (1974) Science 186, 363-364. 157-162. 52. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82, 70-77. Downloaded by guest on September 28, 2021

Top View