Wat. Res. Vol. 21, No. 11, pp. 1339-1344, 1987 Printed in Great Britain. All rights reserved
0043-1354/87 $3.00+0.00 Copyright © 1987 Pergamon Journals Ltd
OXIDATION OF S U L F H Y D R Y L GROUPS BY MONOCHLORAMINE JOSEPH G. JACANGELO*, VINCENT P. OLIVIERI and KAZUYOSHI KAWATA The Johns Hopkins University, School of Hygiene and Public Health, Division of Environmental Health Engineering, 615 North Wolfe Street, Baltimore, MD 21205, U.S.A.
(Received May 1986) Al~tract--In recent years there has been an increased used of monochloramine (NH2C1) for water disinfection because of its low trihalomethane formation potential. Monochloramine is also the predominant disinfectant upon chlorination of wastewater effluents. In an effort to more clearly understand the disinfectant's mode of action in inactivating microorganisms, a study was undertaken to evaluate the compound's reactions with sulfhydryl ( - S H ) groups. The extent of oxidation of these groups was dependent upon the molar ratio of - S H to NH 2CI. When this ratio was > 2: I, the reaction was reversible and ceased at disulfide formation. However, at a ratio of < 2:1, the reaction proceeded irreversibly beyond the disulfide; this reaction continued in the presence of a monochloramine residual. Not all - SH groups in Escherichia coli B were available for reaction. Masking of these _groups within bacterial proteins prevented their complete oxidation at monochloramine doses as high as 100 mg 1-~. The extent to which sulfhydryls are oxidized in bacteria may play an important role in further research on microbial reactivation.
Key words--monochloramine, sulfhydryl group, disulfide, cysteine, cystine, microbial reactivation, oxidation, molar ratio
INTRODUCTION Chlorine has been traditionally employed for the disinfection of water. In recent years, however, there has been increasing concern over the presence of chlorinated organics in finished drinking water. Consequently, much interest has been generated to find an acceptable alternative to chlorine which would keep trihalomethane levels below the maximum contaminant level set by the Environmental Protecton Agency. One such alternative has been the intentional addition of a m m o n i a to chlorine during the water treatment process to form monochloramine. Although the biocidal activity of monochloramine is well documented, little is known about its m o d e of inactivating bacteria during the disinfection process. Studies on compounds containing available chlorine have suggested a reaction of the disinfectant with the amino acid, cysteine, as paramount to the mechanism of cell death (Ingols et al., 1953; Boyle, 1963). Although the chemical reactions of halogens with cysteine have been documented in the literature (Periera, 1973), a paucity of information is available on such reactions with combined chlorine and, in particular, with monochloramine. A previous study (Jacangelo and Olivieri, 1985) showed that monochloramine rapidly reacted with sulfur-containing compounds and tryptophan. Cysteine reacted with m o n o chloramine in a 2:1 stoichiometric relationship (cys*Address for correspondence: 513 West Arlington Place, Chicago, IL 60614, U.S.A.
teine:monochloramine), under conditions in which the disinfectant was the limiting reactant. One mole of the reaction product, cystine, was formed for every mole of monochloramine consumed as shown in the equation below: -2=
NH2CI + 2 R - S H ~ +1
, R - S - S - R + N H 3 + HC! +2e
This reaction is of particular importance because numerous bacterial enzymes contain ( - S H ) groups essential to their active site and the destruction of these groups causes cessation of enzymatic activity. The oxidation of - S H to - S - S is a reversible one under various reducing conditions; however, if the reaction proceeds beyond the disulfide to a greater oxidation state, a reduction back to the original - SH conformation is unlikely. The extent to which the monochloramine-suifhydryl reaction occurs in bacteria is of paramount importance since partial oxidation of the - S H group to a disulfide state could ultimately lead to a bacteriostatic condition rather than to one which precedes cell death. This bacteriostatic condition may be observed as stressed or injured cells. Such a mechanism was proposed by Ingols (1953). Venkobachar et al. (1975) showed partial reactivation of the thiol-dependent enzyme, succinic dehydrogenase, after its activity had been initially inhibited by chlorine and then exposed to glutathione, a reducing compound. Boyle (1963),
JOSEPH G. JACANGELOet al.
to the contrary, reported that sulfhydryl groups were oxidized to elemental sulfur (oxidation state o f 0) and could not be reduced back to its original - S H conformation. The objective o f this study was to determine: (1) under what conditons cysteine oxidation to the disulfide occurs, (2) when the reaction proceeds beyond the disulfide state, and (3) the availability o f sulfhydryl groups in Escherichia coli B for reaction with monocloramine. METHODS
Analyses of sulfhydryl groups Concentrations of cysteine sulhydryl groups ( - S H ) were determined colorimetrically with 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) according to the method of Ellman (1959); modifications were made to adjust for small sample sizes. The DTNB reagent was prepared immediately before use by dissolving with sonication 39.6 mg of the compound in 10 ml of 0.1 M phosphate buffer (pH 7.0). Into 2.0 ml of 0.I M phosphate buffer (pH 8.0), 0.1 ml of sample was added. Immediately afterwards, 0.1 ml of DTNB was introduced into the solution. The absorbance at 412 nm was read after 2 min in a double beam spectrophotometer. The reference cell contained all the reagents except cysteine, which was replaced by twice-distilled, deionized (TDD) water. Standard curves were constructed by plotting absorbance vs moles of sulfhydryl group. Estimations of - S H content in bacteria were made by applying the linear regression equation from the standard curve to the DTNB absorbance values obtained from the assay of bacterial extracts. Dilsulfides were measured using dithioerythritol (DTE) originally described by Cleland (1964) and later refined by Zahler and Cleland (1968). A 0.2 ml sample of the disulfide was mixed with 0.1ml of 0.5 MTris buffer (pH9.0) and 0.1ml of 3.0 × 10 3MDTE. After allowing a reduction time of 20 rain, 1.0 ml of 1.0 M Tris (pH 8.1) and 1.5 ml of 5.0 × 10-3M sodium arsenite were added. After 2min, 0.1 ml of DTNB was added and the absorbance at 412 nm was recorded for 5 min. The absorbance due to the monothiol sulfhydryl groups were determined by extrapolating the linear portion of the curve to the time of addition of DTNB. Controls consisted of samples containing no disulfide. Any absorbance due to controls was subtracted from sample values, which were then compared to a standard curve as described above. Disulfides were reported as moles of - S H . Bacterial preparation and enumeration Escherichia coli B was grown in trypticase yeast extract (TYE) from a previously prepared slant of the culture. The bacteria were incubated at 37°C in a New Brunswick Scientific gyrotory shaker water bath until late log phase, centrifuged at 5000 g for 5 min, and washed three times with 0.01 M phosphate buffer (pH 7.0). After the final washing, the cells were adjusted to the appropriate density by measuring the absorbance in a Spectronic 20 spectrophotometer and comparing to a previously prepared standard curve. The enumeration of E. coli after disinfection experiments was performed by the pour plate method using TYE with 15 g 1- ~ of agar added. Controls consisted of pour plates without bacteria added. Bacterial colonies were counted after 48 h incubation at 37°C using a standard colony counter. Densities of bacteria were reported as colony forming units (CFU) ml-~. Preparation of bacterial extracts Crude cell-free extracts of Escherichia coli B were prepared according to Hughes et al. (1971) using a Virsonic Cell Disrupter Model 16-850. Cells at a density of approx.
5 × 10~cfuml i in 0.01M phosphate buffer (pH 7.0) were chilled to 4°C in an ice bath and then placed in a cold rosette cell. The bacteria were sonicated for 2 min. Plate counts on the crude extract usually yielded <2% of the initial titer of viable cells. Experiments with crude extracts were conducted immediately after preparation. In some experiments sodium laurel sulfate was added to a final concentration of 0.3% in order to denature proteins and further expose sulfhydryl groups.
Monochloramine preparation Stock chlorine solutions were prepared by bubbling 99.5%, high purity chlorine gas through a series of three gas washing bottles. The first bottle contained approx. 100 ml of TDD water and functioned to remove any soluble impurities present in the gas. The second bottle also contained approx. I00 ml of TDD water and was employed as the chlorine stock solution. The third contained double strength nutrient broth and served as a trap for any residual chlorine gas. Monochloramine was prepared by reacting ammonium chloride (NH4C1) with aqueous chlorine at a 3:1 molar ratio (NH4CI to OC1-) at pH 10. Disinfectant concentrations were measured by amperometric titration according to Standard Methods (APHA, 1985) using a Sargent-Welch model XVI polarograph. Monochloramine disinfection system Inactivation experiments were conducted in sterile trypsinizing flasks with magnetic stir bars to assure adequate mixing. Late log phase cells were suspended in 0.01 M phosphate buffer in order to maintain the pH of the system at 7.0. At time zero, the disinfectant was introduced into the flask, the volume of the disinfectant always being approx. 20% of the total volume in the flask. At various times, 8.0 ml samples were removed: 5.0 ml were neutralized with 1.0% by volume of 0.1 M sodium thiosulfate and used for determining survival or performing the biochemical assays; 3.0 ml were used to determine the disinfectant residual. RESULTS
Cysteine oxidation beyond the disulfide Experiments were conducted in order to determine whether the reaction o f m o n o c h l o r a m i n e with cysteine proceeds beyond cystine disulfide to a higher oxidation state. Various molar ratios o f the a m i n o acid to disinfectant were reacted together (pH 7.0 and 25°C) and repetitive u.v. spectral scans o f the reaction were performed over time using a Heath Model E U 700-56 double beam spectrophotometer. W h e n the molar ratios o f cysteine to m o n o c h l o r a m i n e were equal to or exceeded 2: 1, little or no spectral changes were observed over 30 min after the initial reaction. U n d e r such conditions, no disinfectant residuals were detected. Figure 1 shows repetitive scans typical o f the reaction (2:1 molar ratio, cysteine to m o n o chloramine) after 1, 12 and 30 min. A characteristic cystine absorption band with a 250 n m peak was observed. After formation o f the disulfide, no spectral changes were detected, indicating that the reaction has stopped after the formation o f the disulfide. Figure 2 presents repetitive scans o f a similar reaction when the cysteine to m o n o c h l o r a m i n e ratio was < 2:1; in such cases, disinfectant residuals were detected after the reaction. A n a b s o r p t i o n band with a 250 n m peak was initially observed, showing the formation o f the disulfide as an initial reaction
Oxidation of sulfhydryl groups
product. The subsequent degradation of the absorption band was indicative of the reaction proceeding beyond the disulfide to a product in which sulfur has an oxidation state > - 1.
Reversibility of the disulfide The reversibility of the oxidation of cysteine to cystine is presented in Fig. 3. Cysteine and monochloramine were reacted together at different molar ratios at 25°C and pH 7.0. At various times, samples were withdrawn from the reaction flask and incubated with DTE, which reduced disulfides in the reaction mixture to the sulfhydryl state. The presence of a residual was assayed at each sample time. As the molar ratio of cysteine to monochloramine decreased, the reduction of the sulfhydryl from the disulfide also decreased. This figure also demonstrates a trend regarding the extent to which sulfhydryl groups were oxidized. If a disinfectant residual was present, the oxidation continued beyond the disulfide, the latter only being an intermediate in the reaction. At a ratio of 0.23:1.0, cysteine to monochloramine, a residual was detected over the entire contact time, and almost complete loss (94%) of - SH was observed after 60 min. At a 1.0: 1.0 ratio, 72% of - S H was recovered from the disulfide after 15 rain. At longer contact times, however, no residual was detected; further oxidation beyond the disulfide ceased as demonstrated through a tailing-off of the - S H recovery curve. Finally, at a 3.3:1.0 ratio, no residual was present after the initial reaction, and 100% - S H groups were recovered over 30 min; only after 60 min was there a slight loss (8%) in sulfhydryl recovery.
o u 0.4 c 0 JO t_ O 0.3
0.1t 0.0 215
[--ND, 3.3 : 1.0 ~ NH2CL residuol, persistence .-~z / 1.0:1.0 ] 0.6
O. / : 1.0
3.3 : 1.0
\ \ \ \ ~
0.2 K = I ~
\ . 5 rain
10 rain / ~ 3 0 rain
~_ 2o ~ 235
255 275 WoveLength (nm)
0.2 : 1.0 I 15
I 30 Time
I 45 (min)
* I 60
JACANGELO et al.
Table I. Oxidation of sulfhydryl groups in both intact bacteria and bacterial extracts of Escherichia coil B by monochloramine after 5 min contact time at 25'C and pH 7.0. Initial intact bacterial densities wre 6.4 x 10Scfuml
NH2C1 dose (mgl i)
Percent remaining beyond -SH - S - S S-S-.
2 84 10 6 10 59 18 23 100 27 35 38 Bacterial 2 75 21 4 extract 10 29 6(1 1I 100 10 12 78 *100 ND -*Sodium laurel sulfate added to a final concentration of 0.3%. ND~not detected.
Oxidation o f sulfhydryl groups in Escherichia coli B
The oxidation of sulfhydryl groups in both intact bacteria and crude extracts of E. coli B by various doses of monochloramine at 25°C at p H 7 . 0 is presented in Table 1. As the monochloramine dose increased, there was a corresponding increase in oxidation of sulfhydryl groups in both the intact bacteria and the crude extract. However, in intact bacteria, the extent of thiol oxidation was not consistent with bacterial kill. For example, at a dose of 10 mg I 1 of monochloramine, < 50% of thiol groups were oxidized even though 99.9999% of the bacteria were inactivated. Greater oxidation of thiols was observed in the extracts than in the intact bacteria, indicating that some of the - S H groups were not available for reaction in whole cells. When sodium laurel sulfate was added in order to denature proteins and expose previously masked - S H groups, 100% oxidation was observed at 100 mg 1-~ of disinfectant. In intact bacteria, only at the higher doses were appreciable amounts (23 and 38% at 10 and 100 mg 1 ~ of monochloramine, respectively) of sulfhydryl groups oxidized beyond the disulfide state. In the bacterial extract of E. coli, only 4 and 11% - SH was driven beyond the disulfide at 2 and 10 m g l ~, respectively, because of the high disinfectant demand. At 1 0 0 m g l -~ of monochloramine, a residual was present and 78% of the thiol groups were beyond the disulfide state. Figure 4 presents the inactivation of E. coil B and oxidation of the bacterial sulfhydryl groups over time after exposure to 2 . 5 m g l -t of monochloramine. The figure demonstrates that reaction with all - S H groups is unnecessary for inactivation to occur since < 5 0 % of the thiol groups were oxidized after 15min although there was > 7 logs of bacterial kill. DISCUSSION
One of the more important aspects of this study is the clarification of the reaction of monochloramine with sulfhydryl groups. Repetitive spectral scans in Fig. 1 show that when cysteine is in excess (the molar ratio of cysteine to monochloramine is ~>2:1), the reaction appears to stop at the formation of the
NH2CI residual (mgl ~)
Bacterial survival (logN/No)
0.9 7.4 85.4 ND ND 66.8 43.2
- 0.4 - 6.9 >~ 8.7 ---
disulfide. However, if the disinfectant is in excess (the stoichiometry of this ratio becomes <2:1), this reaction progresses beyond this oxidation state. The spectral scans in Fig. 2 show the initial formation of cystine with subsequent degradation of the 250 nm maximum absorbance band. The oxidation of the disulfides can occur by three different pathways (Savige and Maclaren, 1966); various unstable intermediates are involved in this oxidation, the final product being a sulfonic acid. In the case of monochloramine, oxidation probably occurs through a scission of the disulfide to form suifenic and sulfonic intermediates, with cysteic acid as the final reaction product. Periera et al. (1973) recovered cysteic acid when they reacted free chlorine with cysteine. Reactions of other halogens with cystine were also found to yield a cysteic acid reaction product (Gortner and Hoffman, 1927; Shinohara, 1932; Gordon, 1973). Ultimately, the extent to which sulfhydryl groups are
Z -SH control 1.0 0.9
o o t,
E. c o t i
t 10 Time (rain)
Oxidation of sulfhydryl groups oxidized depends upon the presence of a disinfection residual. Figure 3 shows that cysteine oxidation beyond the disulfide continues to occur if a residual is present. However, as soon as it is consumed, the reaction stops at disulfide formation. Although the reaction of monochloramine with cysteine is rapid, thiols in bacteria must be available in order for an oxidation reaction to occur. It is evident from Table 1 that not all - S H groups in E. coli are subject to reaction. In intact bacteria treated with 2 and 10mgl -l, only 16 and 41% of the thiol groups respectively, were oxidized. Further evidence for masked - SH is provided by the experiments with bacterial extracts. Complete thiol oxidation occurred only when the extracts were treated with sodium laurel sulfate, a compound which denatures the macromolecular structure of proteins. The structural complexity of bacterial proteins may also prevent reacted thiol groups from oxidation beyond the disulfide. In the intact bacteria trials, only 6% of the - SH oxidized at 2 mg 1- ~of monocloramine reacted beyond - S - S - , even though a residual was present; at 10 mgl -l of disinfectant, 18% of the - S H oxidized remained at the disulfide. As noted previously, reaction with all - S H groups is unnecessary for bacterial kill. Previous investigators have postulated that only key enzymes were necessary for inactivation (Green and Stumpf, 1946) and in particular, those that were dependent upon sulfhydryl groups (Knox et al., 1948). The unavailability of key, reactive - S H groups to immediate reaction in E. coli may explain why the chemical reactions are much more rapid than the inactivation of the bacteria. The apparent inability of monochloramine to drive the bacterial thiol reaction completely beyond the disulfide in the presence of a residual provides support to the idea that in some cases, a bacteriostatic rather than bactericidal condition may occur. Yet, a reduction back to the - S H conformation is improbable in the presence of a residual. Even at the highest bacterial concentrations normally encountered in water distribution systems (100,000 ml-l), a 0.1 mgl-~ residual would represent a molar ratio of monochloramine to - S H of > 1000:1, respectively. It should be noted, though, that laboratory strains of bacteria grown on rich media do not necessarily reflect the physiological conditions of bacteria grown under low nutrient conditions such as those encountered in distribution systems. The effects of growth conditions have been documented in relation to disinfection sensitivity (Berg et al., 1 9 8 2 , 1983). However, the monochloramine-sulfhydryl studies do emphasize the need to maintain a disinfectant residual. Further research needs to be directed towards defining under which conditions and to what extent bacterial reactivation may occur. Such conditions may exist within biofilms in pipes of water distribution systems where disinfection residuals do not penetrate the film's slime layer. Under reducing con-
ditions, those bacteria in a static state could possibly be reactivated. Such conditions may also exist after chlorinated wastewater effluents are discharged into receiving waters. Resuscitation of monochloraminetreated E. coli was reported by Boyle (1963) after exposure to 3- to 6-fold stoichiometric excesses of sodium thiosulfate, a reducing compound. Similarly, recovery of chlorine-stressed microorganisms has been reported (McFeters and Camper, 1983). Acknowledgements--The authors would like to acknowledge the American Water Works Association Research Foundation and the Long and Widmont Foundation for their support of this work on the mode of action of monochloramine. REFERENCES
APHA (1985) Standard Methods for the Examination of Water and Wastewater, 16th edition. (1985) American Public Health Association, Washington, D.C. Berg J. D., Matin A. and Roberts P. V. (1982) Effect of antecedent growth conditions on sensitivity of E. coli to CIOz. Appl. envir. Microbiol. 44, 814-819. Berg J. D., Matin A. and Roberts P. V. (1983) Growth of disinfectant-resistant bacteria and simulation of natural aquatic environments in the chemostat. In Water Chlorination: Environmental Impact and Health Effects (Edited by Jolley R. L. et al.), Vol. 4, pp. 1137-1147. Ann Arbor Science, Ann Arbor, Mich. Boyle W. C. (1963) Studies on the biochemistry of disinfection by monochloramine. Ph.D. thesis, California Institute of Technology, Pasadena, Calif. Cleland W. W. (1964) Dithiothreitol, a new protective reagent for SH groups. Biochemistry 3, 480-482. Ellman G. (1959) Tissue sulfhydryl groups. Archs Biochem. Biophys. 82, 70--77. Gordon P. G. (1973) The reaction of cystine and alanine-3sulfinicacid with halogens. Aust. J. Chem. 26, 1771-1780. Gortner R. A. and Hoffman W. F. (1927) Sulfur in proteins--III. Derivatives of l-cystine and i-cystine. J. biol. Chem. 72, 433--448. Green D. E. and Stumpf P. K. (1956) The mode of action of chlorine. J. Am. Wat. Wks Ass. 38, 1301-1305. Hughes D. E., Wimpenny J. W. T. and Lloyd D. (1971) Disintegration of microorganisms. In Methods in Microbiology (Edited by Norris J. R. and Ribbons D. W.), Vol. 5B, pp. 1-55. Academic Press, New York. Ingols R. S., Wycoff H. A., Kethley T. W., Hodgden H. W., Fincher E. L., Hildebrand J. C. and Mandel J. E. (1953) Bacterial studies of chlorine. Ind. Engng Chem. 45, 996-1000. Jacangelo J. G. and Olivieri V. P. (1985) Aspects of the mode of action of monochloramine. In Water Chlorination: Chemistry, Environmental Impact and Health Effects (Edited by Jolley R. et al.), Vol. 5, pp. 575-586. Lewis, Chelsea, Mich. Knox W. E., Stumpf P, K., Green D. E. and Auerbach V. H. (1948) The inhibition of sulfhydryl enzymes as the basis of bactericidal action of chlorine. J. Bact. 55, 451-458. McFeters G. A. and Camper A. K. (1983) Enumeration of indicator bacteria exposed to chlorine. Adv. appl. Microbiol. 29, 177-193. Pereira W., Hoyano Y., Summons R., Bacon V. and Duffield A. (1973) Chlorination studies---II. The reaction of aqueous hypochlorous acid with a-amino acids and depeptides. Biochim. biophys. Acta 313, 170-180. Savige W. E. and Maclaren J. A. (1966) Oxidation of disulfides, with special reference to cystine. In The Chemistry of Organic Sulfur Compounds (Edited by Karasch N.
JOSEPH G. JACANGELOet al.
and Meyers C. Y.), Vol. 2, pp. 367-402. Pergamon Press, Oxford. Shinohara K. (1932) Oxidation of cystine by iodine in aqueous medium. J. biol. Chem. 96, 285-297.
Venkobachar C., lyengar L. and Rao A. V. S. P. (1975) Mechanism of disinfection. Wat. Res. 9, 119-124. Zahler W. L. and Cleland W. W. (1968) A specific and sensitive assay for disulfides. J. biol. Chem. 143, 716-719.