Journal of Applied Bacteriology 1990.68.453459
3 1 10/06/89
Inactivation of Legionella pneumophila by monochloramine D. A . C U N L I F FState E Water Laboratory, Engineering and Water Supply Department, Salisbury, South Australia, 5108 Accepted 31 July 1989
C U N L I F F D. E , A . 1990. Inactivation of Legionella pneumophila by monochloramine. Journal of Applied Bacteriology 68,453-459. Chloramination which is used in South Australia to control the growth of Naegleria fowleri, was investigated to see if it would also control that of Legionella pneumophila. It was found that L. pneumophila was more sensitive than Escherichia coli to monochloramine. At 1.0 mg/l, a 99% kill of L. pneumophila was achieved in I5 min compared with 37 min for a 99% kill of E . coli. Combined with the stability of monochloramine, even at elevated temperatures, the results suggest that this disinfectant would control the growth of L. pneumophila in water distribution systems.
Legionella pneumophila has been isolated from a wide range of environments including water, sediments and biofilms from water distribution systems (Wadowsky et al. 1982; Bartlett 1983; Ciesielski et al. 1984; Colbourne ef a/. 1984; Stout et a / . 1985; Colbourne et al. 1988). The presence of L. pneumophila in plumbing systems of large buildings has been associated with outbreaks of Legionnaires disease (Stout et al. 1982; Bartlett 1983; Bartlett et a/. 1983; Shands et a/. 1985). In the majority of cases the building received its water from a public water supply. After outbreaks, either heat or chlorine have been used to control the spread of the organism (Baird et a/. 1983; Bartlett 1983; Bartlett ef al. 1983; Best et a/. 1983; Massanari ef a/. 1983). In addition, although there are some disadvantages associated with rapid decay and corrosion (Fraser 1985; Muraca et al. 1987, 1988). some hospitals have introduced free residual chlorination as a long-term control measure. In South Australia residual chlorination has been replaced with chloramination in several public supplies where it was necessary to control Naegleria fowleri, the causative agent of primary amoebic meningoencephalitis (Carter 1970). The principal advantage of monochloramine is its relative stability and it has been general practice in South Australia to maintain a concentration of 1 G 1 . 5 mg/l mono-
chloramine throughout the public distribution systems. The persistence of monochloramine, however, presents a problem to a hospital that adds chlorine to the plumbing system, as mixing of the two has to be carefully controlled to overcome the demand and to generate a free residual chlorine (Barrett et a / . 1985). An alternative would be to rely on monochloramine present in the public supply to control the growth of L. pneumophila in the plumbing system of the hospital. This work examined the relative sensitivities of L. pneumophila and Escherichia coli to inactivation by monochloramine.
Materials and Methods B A C T E R I A A N D CULTURE M E D I A
Legionella pneumophila serogroup 1 was obtained from the Institute of Medical and Veterinary Science, Adelaide, and was maintained on slopes of Buffered Charcoal Yeast Extract (BCYE) agar (Oxoid). In the disinfection trials, it was enumerated by plate counts on BCYE agar plates. The E . coli was an environmental strain isolated at the State Water Laboratory and was enumerated by plate counts on Tryptone Soya Agar (Oxoid).
454 MONOCHLORAMINE S O L U T I O N
Monochloramine was prepared by mixing ammonium chloride with sodium hypochlorite in a 2.5 : 1 molar ratio. Stock monochloramine at 5W600 mgA was prepared by dissolving 1.15 g of ammonium chloride in 900 ml of distilled water, adjusting the pH to 8.5-9.0 with sodium hydroxide and then slowly adding 6 ml of sodium hypochlorite (10% w/v CI,; 600 mg CI,). The final volume was made up to 1 1. Stock monochloramine was stored in the dark at 4°C.
Monochloramine concentrations were determined by titration with ferrous ammonium sulphate with N,N-diethyl-p-phenylenediamineas a colorimetric indicator (Anon. 1985).
A simple plumbing system was used as a reac-
tion vessel (Fig. 1). Several components of this system were derived from the more complex model described by Muraca et al. (1987). The system included a 20 1 glass reservoir, a 2 I glass jar containing a mixture of rubber washers to a depth of 40 mm, 7-5 m of 9 mm diam. copper tubing, 4 m of 10 mm diam. silicone tubing and
Cunlife a variable speed peristaltic pump. The two glass containers were held in a water bath. D I S I N F E C T I O N EXPERIMENTS
Before each experiment the plumbing system was flushed and then filled with 18 I of hot (8CL 90°C) autoclaved tap water at pH between 8.4 and 8.6. The water was allowed to cool to 30°C and then inoculated with either L. pneumophila o r E. coli. Legionella pneumophila was grown on slopes of BCYE agar which were incubated for 72 h at 37°C. Bacterial suspensions were prepared by washing three slopes with sterile tap water. This provided an initial concentration of 3 4 x lo5 cells/ml. The L. pneumophila was then allowed to circulate in the plumbing system at a flow rate of about 1 I/min for 2 h. Immediately before adding a predetermined volume of stock monochloramine t o the 20 I container, a 1 ml sample was collected and the flow rate was increased to 3 4 I/min to facilitate mixing of the disinfectant. After the addition of monochloramine, further 1 ml samples were taken at 2 min intervals for 20 min and then at 25, 30 and 40 min. Each sample was serially diluted in sterile tap water. Each dilution bottle contained 1 ml of 10% (w/v) sodium thiosulphate, to neutralize monochloramine, and 0.1 ml volumes of each dilution were plated on BCYE agar in triplicate. The plates were incubated at 37°C in air containing
I glass vessel
2 I glass vessel Fig. 1. Schematic diagram of the model plumbing system used in the disinfection experiments.
Znactioation of L. pneumophila 3% CO,. Colonies of L. pneumophila were counted after incubation for 72 h. Monochloramine concentrations were determined after 2 and 40 min. The loss of monochloramine never exceeded 0.1 mg/l. Experiments with E. coli were similar but the organisms were harvested by centrifugation (7000 g for 15 min) of 10 ml of an overnight Tryptone Soya Broth (Oxoid)culture, incubated at 35°C for 18 h. The bacteria were resuspended in 10 ml of sterile tap water, inoculated into the plumbing system and allowed to circulate for 30 min before the addition of monochloramine. After the addition of monochloramine, 1 ml samples were collected at 2 rnin intervals for 20 rnin and then at 5 min intervals for a further 40 min. The initial concentration of E. coli in the system was 3 4 x lo5 cells/ml. Colonies of E . coli were counted after incubation at 35°C for 24 h. DETERMINATION OF
I N A C T I V A T I O N TIMES
For each disinfection experiment, the time taken to inactivate 99% of the original inoculum was determined by plotting log percentage survival against time (min). After initial lag periods, these plots produced straight lines with correlation coeflicients (r'), as determined by regression analysis, in the range 0.96-0.99.
S T A B I L I T Y OF M O N O C H L O R A M I N E
Samples of potable water containing monochloramine were collected from public supplies and were incubated in closed flasks at either 30°C or 55°C. Monochloramine concentrations were monitored throughout the course of the incubation.
Results The ability of L. pneumophila to survive in the model system in the absence of monochloramine is shown in Fig. 2. There is no significant reduction in the numbers of viable organisms over a 6 day period. The results of the disinfection experiments show that L. pneumophila is more sensitive than E . coli to inactivation by monochloramine (Fig. 3). The results are expressed in terms of the formula of Watson (1908) in which k = C" x t where k is a constant, C is the concentration of disinfectant, t is the time taken to inactivate 99% (in this case) of the original inoculum and n, the slope of the lines in Fig. 3, is the coeflicient of dilution. Comparisons of C x t products for different organisms require n to approximate 1.0 (Rubin et al. 1983; Wickramanayake et nl. 1984). The values of n from Fig. 3 were 0.98 for L. pneumophila and 0.97 for E. coli. The average C x t,, for L. pneumophila
Time ( d )
Fig. 2. Survival of Legionello pneumophila in the model system. The system was prepared, inoculated with L. pneumophila and sampled as described for the disinfection experiments. Each point represents the average of triplicate analyses.
was 15 mg.min/l while for E. coli it was 37 mg.min/l. In the disinfection experiments with L. pneumophila the bacteria were inoculated into the model system 2 h before the addition of monochloramine. Extending the time between inoculation and addition of monochloramine to up to 6 d did not increase the resistance of L. pneumophila to inactivation by monochloramine (results not shown). Colbourne et a / . (1988) reported that heat shock can lead to the recovery of previously 'viable non-culturable' L. pneumophila from water samples. This technique was tried but heat shock did not improve the recoveries of L. pneumophila during disinfection experiments. The decay of monochloramine in potable water was measured at 30°C and 55°C. Before incubation the water contained 1.3 mg/l monochloramine. As expected, the decay was more rapid a t 55°C but even after 50 h the water still contained 0.35 mg/l (Fig. 4). After 5 d a t 30°C the water contained 0.8 mg/l (Fig. 4).
Discussion Experimental conditions were chosen both to favour the survival of L. pneumophila and to be consistent with those found in local chloraminated supplies. This was achieved by a simple plumbing system and tap water at 30°C and pH 84-8.6. In South Australia, chloramination has generally been introduced for supplies incorporating long above-ground pipelines and, in summer, water temperatures are typically in the range 25-30°C. Where possible relatively high pH levels are maintained in these supplies to promote the stability of monochloramine. The disinfection experiments showed that I.. pneumophila was more sensitive to monochloramine than the faecal indicator E. coli. The average C x t , , for L. pneumophila was 15 mg.min/l compared with 37 mg.min/l for E. coli. In comparison it has been found that I,. pneumophila is less sensitive than E. coli to disinfection by chlorine (Kuchta et a / . 1983). There are two advantages in using monochloramine as a biocide for L. pneumophila in
Inactivation of L. pneumophila
12 I I
-& 0 9 -E 2
!? 0 7
c 06 -
04 0 3
Time ( h 1
Fig. 4. Decay of monochloramine in potable water incubated at I, 30°C and A, 55°C. The water used was collected from a chloraminated public supply. Each point represents the average of triplicate analysers.
plumbing systems: it is relatively stable; and it can penetrate biofilms (LeChevallier et a / . 1988). Legionella pneumophila has been found in water and sediments from hot water systems (FisherHoch et al. 1982; Wadowsky et al. 1982; Stout et al. 1985) and from the surfaces of pipes and plumbing fittings (Ciesielski el a/. 1984; Colbourne et a/. 1984; Voss et a/. 1986). The colonization of hot water systems by L.pneumophila has been a particular problem in hospitals (Fisher-Hoch et a / . 1982; Wadowsky et a / . 1982; Stout et a/. 1985). Control of L. pneumophila in these systems requires the maintenance of tempertures 60°C or greater in both the storage tank and distribution mains (Groothuis et a/. 1985; Stout et a/. 1986). However, hospital water systems are often kept below this temperature to reduce the risk of scalding. Even at elevated temperatures monochloramine is relatively stable. Incubation of chloraminated water
at 55°C resulted in only a gradual loss of disinfectant; after 50 h 0.35 mg/l remained in samples that originally contained I .3 mg/l monochloramine. The greater stability of monochloramine at lower temperatures combined with its ability to penetrate biofilms should control the levels of L. pneurnophila in cooler parts of plumbing systems. In laboratory experiments alternative disinfectants such as chlorine, heat, ozone and U.V. light have been found to be effective against L. pneumophila (Muraca et a / . 1987). In practice, however, each has limitations when used to provide continual protection throughout a water distribution system (Muraca ei al. 1988). Heat eradication is generally restricted to use as a shock treatment, U.V. light and ozone d o not offer any residual protection and chlorine decays rapidly. Extrapolating results from laboratory experi-
D . A . Cunlifle
ments to field situations can be dificult. For example, it has been reported that agarpassaged L. pneumophila is more sensitive to chlorine than non-passaged L. pneumophila (Kuchta et al. 1985). Conversely, it has been observed that laboratory results underestimate the effectiveness of monochloramine in field situations (Wolfe & Olson 1985). On balance, however, it is concluded that the presence of monochloramine in water should contribute to the prevention of growth of L. pneumophila in plumbing systems in large buildings. In addition, monochloramine may be useful as a biocide for L. pneumophila in water distribution systems and cooling towers.
evaluation of a method for control of nosocomial Legionnaires disease and Pittsburgh pneumonia. Lancet ii, 307-310. CARTER, R.F. 1970 Description of a Naegleria isolated from two cases of primary amebic encephalitis, and of the experimental pathological changes induced by it. Journal of Pathology 100,217-244. CIFSIELSKI,C.A., BLASER, M.J. & WANG,W.L. 1984 Role of stagnation and obstruction of water flow in isolation of Legionella pneumophila from hospital plumbing. Applied and Enoironmental Microbiology 48,984987. COLBOURNE, J.S., PRATT, D.J., SMITH,M.G., FISHERHOCH, S.P. & HARPER,D. 1984 Water fittings as sources of Legionella pneumophila in a hospital plumbing system. Lancer i, 21G213. COLBOURNE, J.S., DENNIS,P.J., TREW,R.M., BERRY,C. & VESEY, G . 1988 Legionella and public water supplies. Water Science and Technology 20 (11/12). 5- 10. I thank P. Thomas for the provision of monoFISHER-HOCH, S.P., SMITH,M.G. & COLBOURNE, J.S. chloramine solutions and for assistance in deter1982 Logionella pneumophila in hospital hot water mining monochloramine concentrations and G. cylinders. Lancet i, 1073. FRASER,D.W. 1985 Potable water as a source for Wood, P. Christy and R.P. Walters for helpful legionellosis. Enoironmental Health Perspectives 62, discussions. 337-341. This research was supported in part by a GROOTHUIS, D.G., VEENENDAAL, H.R. & DIJKSTRA, grant from the Urban Water Research AssociH.L. 1985 Influence of temperature on the number of Legionella pneumophila in hot water systems. ation of Australia. Journal of Applied Bacteriology 59, 529-536. KUCHTA,J.M., STATES, S.J., MCNAMARA,A.M., WAWWSKY, R.M. & YEE, R.B. 1983 Susceptibility References of Legionella pneumophila to chlorine in tap water. ANON.1985 Standard Methods for the Examination of Applied and Enoironmental Microbiology 46, 1 1 3 4 Water and Wastewater 16th edn, pp. 306-309. 1139. Washington: American Public Health Association. KUCHTA, J.M., STATES,S.J., MCLAUGHLIN, J.E., OVERBAIRD,I.M.. POTTS, W., SMILEY, J., CLICK, N., SCHLEMEYER, J.E., WADOWSKY, R.M., MCNAMARA, A.M., ICH, S., ~ N N O L C. E , & DAVISON, K. 1983 Control WOLMRD,R.S. & YEE,R.B. 1985. Enhanced chloof endemic nosocomial legionellosis by hyperrine resistance of tap water-adapted Legionella chlorination of potable water. In Logionella: Propneumophila as compared with agar mediumceedings of the Second International Symposium ed. passaged strains. Applied and Enoironmental MicroThornsberry C., Balows. A., Feeley, J.C. & Jakubiology 50. 21-26. bowski, W. p. 333. Washington: American Society LECHEVALLIER. M.W., CAWTHON, C.D. & LEE, R.G. for Microbiology. 1988 Inactivation of biofilm bacteria. Applied and BARREIT, S.E., DAVIS, M.K.& MCGUIRE,M.J. 1985 Environmental Microbiology 54,2492-2499, Blending chloraminated and chlorinated waters. MASSANARI, R.M., HELMS,C., ZEITLER, R., STREED,S., Journal of the American Water Works Association GILCHRIST,M., HALL, N.. HAUSLERJR., W., 17. 5 M 1 . JOHNSON, W., WINTERMEYER, L., MUH$ J.S. & HIERBARTLETT,C.L.R. 1983 Potable water as reservoir and HOLZER JR., W.J. 1983 Continuous hypermeans of transmission. In Logionella: Proceedings chlorination of a potable water system for control of the Second International Symposium ed. Thornof nosocomial Logionella pneumophila infections. In sberry, C., Balows, A., Feeley, J.C. & Jakubowski, Legionella: Proceedings of the Second International W. pp. 21Cb215.Washington: American Society for Symposium ed. Thornsberry, C., Balows, A., Feeley, Microbiology. J.C. & Jakubowski, W. pp. 334-336. Washington: BARTLETT,C.L.R., SWANN,R.A., CASAL,J., CANADA American Society for Microbiology. Rouo, L. & TAYLOR,A.G. 1983 Recurrent LegionMURACA, P., STOUT,J.E. & Yu, V.L. 1987 Comparanaires disease from a hotel water system. In Legiontive assessment of chlorine, heat, ozone and UV ella: Proceedings of the Second International light for killing Legionella pneurnophila within a Symposium ed. Thornsberry, C., Balows, A., Feeley, model plumbing system. Applied and Enoironmental J.C. & Jakubowski, W. pp. 237-239. Washington: Microbiology 53,447-453. American Society from Microbiology. MURACA, P.W., YU, V.L. & STOUT,J.E. 1988 EnvironBFST, M., Yu, V.L., STOUT,J., GOETZ,A., MUDER,R.R. mental aspects of Legionnaires disease. Journal of & TAYLOR, F. 1983 Legionellaceae in the hospital the American Water Works Association 80.78-86. water supply epidemiological link with disease and RUBIN,A.J., ENGEL, J.P. & SPROUL,O.J. 1983 Disin-
Inactivation of L. pneumophila fection of amoebic cysts in water with free chlorine. Journal Water Pollution Control Federation 55, 1174-1182. SHANDS, K.N., Ho, J.L., M m , R.D., GORMAN, G.W., EDELSTEIN, P.H., MALLISON, G.F., FINEGOLD,S.M. & FRASER,D.W. 1985 Potable water as a source of Legionnaires disease. Journal of the American Medical Association 256, 1412-1416. STOUT,J.. Yu, V.L., VICKERS, R.M., ZURAVLEFF, J., BEST, M., BROWN,A., YEE, R.B. & WAWWSKY, R.M. 1982. Ubiquitousness of Legionella pneumophila in the water supply of a hospital with endemic Legionnaires disease. New England Journal of Medicine 306,466-468. STOUT,J.E., Yu, V.L. & BEST,M.G. 1985 Ecology of Legionella pneumophila within water distribution systems. Applied and Enoironmental Microbiology 49,221-228. STOUT,J.E., BEST,G. & Yu, V.L. 1986 Susceptibility of members of the family Legionellaceae to thermal stress: Implications for heat eradication methods in water distribution systems. Applied and Enoironmental Microbiology 52, 396399.
WAWWSKY, R.M., YEF, R.B., MEZMAR,L., WING,E.J. & DOWLING, J.N. 1982 Hot water systems as sources of Legionella pneumophila in hospital and non-hospital plumbing fixtures. Applied and Enoironmental Microbiology 43, 1104-1 110. WATSON,H.E.1908 A note on the variation of the rate of disinfection with change in the concentration of the disinfectant. Journal of Hygiene 8,536-542. WICKRAMANAYAKE. G.B., RUBIN,A.J. & SPROUL,O.J. 1984 Inactivation of Naegleria and Giardia cysts in water by ozonation. Journal Water Pollution Control Federation 56,983-988. WOLW R.L. & OLSON,B.H.1985 Inability of laboratory models to accurately predict field performance of disinfectants. Water Chlorination: Chemistry, Enoironmental Impact and Health Effects Vol. 5, ed. Jolley, R.L., Bull, R.J., Davis, W.P., Katz, S., Roberts, M.H. & Jacobs, V.A. pp. 555-573. Michigan: Lewis Publishers. V m , L., BUTTON,K.S.,LORENZ,R.C. & TUOVINEN, O.H. 1986 Logionella contamination of a preoperational treatment plant. Journal of the American Water Works Association 78, 7C75.