Inactivation of Giardia muris Cysts by Free Chlorine

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, JUIY 1987, p /87/ $02.00O0 Copyright 1987, American Society for Microbiology Vol. 53, No. 7 Inactivation of Giardia muris Cysts by Free Chlorine JOSEPH G. LEAHY,t ALAN J. RUBIN,t* AND OTIS J. SPROUL Water Resources Center, Ohio State University, Columbus, Ohio Received 4 August 1986/Accepted 23 March 1987 The chlorine resistance of cysts of the flagellate protooan Giardia muris was examined. This organism, which is pathogenic to mice, is being considered as a model for the inactivation of the human pathogen Giardia lamblia. Excystation was used as the criterion for cyst viability. Experiments were performed at ph 5, 7, and 9 at 25 C and ph 7 at 5C. Survival curves were "stepladder"-shaped, but concentration-time data generally conformed to Watson's Law. Chlorine was most effective at neutral ph and was only slightly less so in acidic solutions. Comparison of inactivation data based on equivalent hypochlorous acid concentrations, which corrects for chlorine ioniation, showed that the cysts have a ph-dependent resistance to inactivation. Concentration-time (C. t') products for free chlorine obtained at 25 C ranged from a low of 50 mg min/liter at ph 5 to a high of 218 mg min/liter at ph 9 and were as high as 1,000 mg min/liter at 5 C. It appears that G. muris cysts are somewhat more resistant to inactivation than G. lamblia cysts and rank among the microorganisms that are most resistant to inactivation by free chlorine. The parasitic flagellate Giardia lamblia is the causative agent of giardiasis, a severe diarrheal disease of humans (8). Endemic to nearly every region of the world (16), this protooan has been implicated in increasing numbers of outbreaks of the disease in the United States over the last two decades (4-9). Contamination of potable water supplies with G. lamblia cysts has been established as a source of many of the outbreaks (13). The failure of water treatment practices to remove or inactivate the cysts has occurred in systems in which filtration is not employed, or in systems in which plain chlorination is the principal form of treatment (6, 13). This suggests that the cyst form of the organism may be resistant to destruction by chlorine. Few studies of Giardia inactivation by chlorine have been reported in the literature. Jarroll et al. (14) found that G. lamblia cysts were resistant to chlorine at elevated ph or low temperature but could be destroyed with moderate chlorine residuals if temperatures were sufficiently high. At 5 C, no cysts survived a 10-min exposure to 8 mg of chlorine per liter at ph 6 and 7 or a 30-min exposure at ph 8. At 25 C, a residual of only 1.5 mg of chlorine per liter inactivated all of the cysts in 10 min at ph 6, 7, and 8. A similar study was performed by Rice et al. (17) with cysts of G. lamblia from symptomatic and asymptomatic donors and with cysts of Giardia muris, a parasite of mice. By employing 2.5 mg of chlorine per liter at 5 C, greater than 99% inactivation of the G. lamblia cysts from the symptomatic donor was obtained in 30 min at ph 6 and in 60 min at ph 7 and 8. Under the same conditions, only 90 to 99% inactivation of G. muris or G. lamblia cysts from the asymptomatic donor was possible. Wickramanayake et al. (27) have compared the inactivation of G. lamblia and G. muris by oone. They found that G. muris is slightly more resistant than the human parasite and proposed that G. muris be used as a model for the inactivation of G. lamblia. * Corresponding author. t Present address: Department of Microbiology, University of Maryland, College Park, MD t Present address: Department of Civil Engineering, The Ohio State University, Columbus, OH Present address: College of Engineering and Physical Sciences, University of New Hampshire, Durham, NH The intent of this study was to provide additional data on the extent and kinetics of the inactivation of G. muris cysts by free chlorine. The effects of temperature and ph were examined at 5 and 25 C and at ph 5, 7, and 9. Concentrationtime (C. t') products and values for the coefficient of dilution were calculated from the data, as described in previous studies (20, 27). MATERIALS AND METHODS Cyst preparation and enumeration. G. muris cysts were maintained by the method of Roberts-Thomson et al. (19) with minor modifications. The hosts were 21- to 28-day-old nonimmune, outbred female Swiss albino mice (CF-1). The mice were infected by inoculation of a suspension of 5,000 G. muris cysts per ml into the lower esophagus; fresh stool samples were collected 6 to 14 days later. Cysts were isolated from the stool samples and suspended in 0.1% Tween 20 detergent solution (19). The cyst concentration of the suspensions was determined by direct count with a hemacytometer. Viable cysts were quantitated by the excystation procedure described by Schaeffer et al. (22). Trophooites were induced to excyst during incubation at 35 C in a combination bicarbonate-reducing solution and later in a solution of glucose-inorganic salts-trypsin. The resulting suspension of cysts and excysted trophooites was placed in the well of a depression slide, which was examined by using an inverted phase-contrast microscope. Empty cyst walls and partially excysted trophooites were counted as positive excystations. Intact cysts were counted as being nonviable. The precision of the technique, +3% for the same batch of cysts, was satisfactory for inactivation experiments. Solutions and analyses. All solutions were prepared with demand-free water as follows. Deionied water from a reverse-osmosis, ion-exchange system (Milli-Qi Millipore Corp., Bedford, Mass.) was dosed with sodium hypochlorite solution to yield a residual of 5 mg of free chlorine per liter. After standing protected from sunlight for 24 h, the remaining chlorine residual was removed by irradiating the solution with an immersible UV lamp. Stock buffer solutions for the cyst inactivation studies were prepared so that the final diluted total concentrations were 2.5 x 10-3 M. The ph 5 buffer was 0.18 M in acetic acid

2 VOL. 53, 1987 INACTIVATION OF GIARDIA CYSTS BY CHLORINE o 5 10 s5 20 CONTACT TIME, MIN. FIG. 1. Inactivation of G. muris cysts by free chlorine at ph 7 and 25 C. Residual free chlorine concentrations were 2.80 mg/liter (U), 5.20 mg/liter (0), and 7.12 mg/liter (K). and 0.32 M in sodium acetate, the ph 7 buffer was 0.28 M in potassium dihydrogen phosphate and 0.22 M in dipotassium hydrogen phosphate, and the ph 9 buffer was 0.48 M in sodium bicarbonate and 0.22 M in sodium carbonate. If necessary, the final ph was adjusted prior to an experiment with a few drops of either 1.2 M HCl or 1.0 M NaOH. In none of the experiments did the ph change by more than ±0.04 units, as measured with a ph meter (model NX; Sargent-Welch) and combination electrode. Free chlorine stock solutions of approximately 5,000 mg/liter were prepared by diluting a commercial sodium hypochlorite solution (Clorox) with chlorine demand-free water. Free and combined chlorine were determined by amperometric titration with M phenylarsine oxide solution (1). Inactivation procedures. One-liter amber borosilicate bottles fitted with thermometers and air supply ports were used as reaction vessels. During experimentation test and control reactors were pressuried slightly with compressed air to allow for easy sample collection. Magnetic stirrers were employed for continuous mixing. Temperature control for the 25 C experiments was accomplished by immersing the bottles in a constant-temperature bath. The bottles for the 5 C experiments were placed in water baths located in a thermostatically controlled room. The following was the general cyst-inactivation procedure. A solution of 750 ml containing buffer and free chlorine was placed in the test reactor. A separate reactor containing only buffer served as the control. Both reactors were immersed in the water bath, and initial samples were removed for ph and free chlorine measurements. Aqueous 3-ml suspensions of 1- to 7-day-old cysts were then introduced simultaneously into the reactors with hypodermic syringes, yielding concentrations of approximately 10,000 cysts/ml. Samples of 15 ml were removed at suitable intervals and transferred into screw-cap tubes containing 1 ml of 0.05 M sodium thiosulfate. A minimum of three samples were taken from the control reactor. Free chlorine residuals and ph were measured at the end of an experiment and at appropriate intervals during experimentation. Temperature was monitored closely and maintained within +0.5 C of the desired value. The samples were centrifuged following completion of an experiment, and all but 0.5 ml of the supernatant was removed. Cyst viability was then determined by selected samples by the methods described above. The percent excystation for test samples was calculated relative to that of a control. RESULTS AND DISCUSSION Experiments were conducted to determine the rate of inactivation of G. muris cysts by free chlorine at 5 C at ph 7 and 25 C at ph 5, 7, and 9. Survival curves were obtained for these experiments by plotting the logarithm of the ratio of live cysts to total cysts against contact time. Cyst inactivation at ph 7 by free chlorine at 25 C and 5 C is shown in Fig. 1 and 2, respectively. These curves are representative of all of the survival curves in their "stepladder" shape. The initial portion of the typical curve consisted of a period of rapid cyst inactivation to a level that varied from 5 to 70% survival, depending on chlorine concentration, ph, and temperature. A period characteried by a markedly lower rate of cyst inactivation ensued as the curve developed a shoulder corresponding to a lag period of inactivation. The remainder of the curve followed with a rapid decrease in cyst survival. Finally, the inactivation rate appeared to be constant as the curve became linear. The time required to obtain 99% kill was determined by interpolation between the data points with the lowest percentage survival above 1% and the highest percentage survival below 1%. The experimental results are presented in Table 1. The HOCI dissociation constants used for the calculations were 1.7 x 10-8 for the experiments at 5 C and 2.7 x 10-8 for those at 25 C (21). There were no significant losses of free chlorine during the experiments, as measured by the amperometric titration method. The contact time for a given ph and temperature required to yield 99% inactivation increased as the free chlorine residuals decreased. The multiple of the time required to effect a given kill and the corresponding disinfectant concentration provide a convenient means for comparing inactivation data at any specified temperature and ph. This C- t' product is related to the Watson's law equation: Cat' = k, where C is the residual chlorine concentration, n is a constant known as the coefficient of dilution, t' is the contact time required for a fixed percent inactivation, and k is a constant (20). The C t' products for 99% cyst inactivation by total free chlorine and HOCI and the values of the coefficients of dilution found in this study are summaried in Table 1. -1I -2-3 a CONTRCT TIME. MIN. FIG. 2. Inactivation of G. muris cysts by free chlorine at ph 7 and 5 C. Residual free chlorine concentrations were 31.4 mg/liter (H), 52.5 mg/liter (0), and 78.5 mg/liter (O).

3 1450 LEAHY ET AL. APPL. ENVIRON. MICROBIOL. TABLE 1. Inactivation of G. muris cysts by free chlorine Concn (mg/liter) (mg C min/liter) t' product for: Coefficient Temp ph 99% kill time of dilution (OC) Total chlorine Residual (min) Total HOCI o ) Initial Residual HOCI chlorine (0.989) (0.988) (0.994) , (0.987) a r, Correlation coefficient. The coefficient of dilution from the Watson's law equation is indicative of the relative importance of disinfectant concentration and contact time in the disinfection process. Values of n larger than 1 indicate that the effectiveness of a disinfectant is rapidly diminished as it is diluted. When values are less than 1 the contact time assumes greater importance. In this study a coefficient of dilution near 1 (0.93) was obtained only at ph 9 and 25 C. Under all other conditions the value of n was significantly greater than 1, indicating that the concentration of free chlorine or HOCI is more important than contact time in the inactivation of G. muris cysts. Watson's Law predicts a linear fit when the logarithm of disinfectant concentration is plotted against the logarithm of the time necessary to achieve a given percentage kill. Regression by least-squares analysis of log C-log t' data for the inactivation of G. muris cysts for free chlorine yielded correlation coefficients ranging from to These values imply a high correlation between log C and log t' and therefore strong conformity to Watson's Law. Such plots of the data in terms of free chlorine and HOCI residuals are given in Fig. 3 and 4, respectively. The points are experimental data and the straight lines were determined by least squares regression. Temperature had an important effect on the results, as indicated by the results obtained at ph 7 (Fig. 3). The effectiveness of free chlorine at 25 C was about 10 times greater than that at 5 C. Much larger C. t' products were required at 5 C than at 25 C, on the basis of their calculation from both free chlorine and HOCI residuals. The effect of temperature has a greater significance with cysts than with bacteria because the rate of inactivation is so much greater for the latter. The effect of ph on the efficacy of both free chlorine and HOCI against G. muris cysts was also significant. The largest 20 IC ph5 E 0 0 E '2 <F w u 0 ph 7 99% KILL TIME, min, FIG. 3. Watson's Law plots for the inactivation of G. muris cysts with total free chlorine at ph 5 (*), ph 7 (O), and ph 9 (0) at 25 C and at ph 7 at 5 C (U) % KILL TIME, min. FIG. 4. Watson's Law plots for the inactivation of G. muris cysts with hypochlorous acid (calculated from total free chlorine) at ph 5 (*), ph 7 (O), and ph 9 (0) at 25 C. 50

4 VOL. 53, 1987 INACTIVATION OF GIARDIA CYSTS BY CHLORINE 1451 TABLE 2. Comparison of C * t' products for the 99% inactivation of different organisms by free chlorine Test organism ph Temp (OC) Concn (mg/liter) Cm/t' (mg min/liter) Reference Protooan cysts G. muris , G. lamblia <isa a 17 N. gruberi N. fowleri A. culbertsoni ,500-7,200a 10 A. sp. strain 4A ,000a 10 E. histolytica a 24 Viruses and bacteria Poliovirus (Mahoney) a a 11 Poliovirus 1 (Brunhilde) a 23 Poliovirus a 11 Echovirus a 11 Echovirus a Sa 28 Coxsachievirus A a 11 Coxsachievirus B a a a 15 Coxsackievirus B a 15 E. coli a a Values were approximated from data in the original report. C- t' products for the inactivation of G. muris cysts were required at ph 9, with the smallest values being required at ph 7 when free chlorine residuals were used to calculate the values. In terms of HOCI residuals, the C- t' products required for 99% inactivation decreased with increasing ph. Total free chlorine was more effective at ph 7 than at ph 5 and was much less effective at ph 9 (Fig. 3). In contrast, the same data plotted in terms of HOCI concentrations show that the effectiveness of HOCI in the inactivation of G. muris cysts is almost directly related to ph, with smaller residuals being required to achieve 99% inactivation as the ph increased from 5 to 9. The irregular effect of ph on the inactivation of G. muris cysts by free chlorine was unusual and requires comment. It has been established in previous studies that free chlorine becomes less effective as the solution ph is raised. This variation in effectiveness is due to the ph-dependent distribution of chlorine between hypochlorite ion and hypochlorous acid, with the latter being the active disinfecting agent in free chlorine solutions. For a dissociation constant of 2.7 x 10-8, HOCI constitutes 99.7% of the total free chlorine at ph 5 and 78.7% at ph 7, and is only 3.6% of the total residual concentration at ph 9. Therefore, in contrast to the present results, free chlorine would be expected to be more effective at ph 5 than at ph 7. It is generally assumed that a given microorganism is equally sensitive to HOCI over small ranges of ph. As described by Rubin et al. (20), this is the case with another protooan, Naegleria gruberi. In a study of the inactivation of N. gruberi cysts at 25 C with free chlorine expressed as HOCI, there was little significant difference found between plots of log C-log t' data at ph 5, 7, and 9. The phenomenon of increasing effectiveness of HOCI with increasing ph has also been reported for other organisms, for example, for poliovirus by Englebrecht et al. (11) and by Taylor and Butler (25). The ph change causes alterations to the surface of the organisms. This may effect their sensitivity to HOCl or, alternately, the mode of inactivation of the disinfectant may be different, leading to enhanced sensitivity to OCl-, which is increased with ph at any fixed HOCI concentration. The magnitude of the free chlorine residuals required to inactivate G. muris cysts demonstrates the resistance of this organism to chlorine. Residuals of about 3 and 4 mg/liter are needed to inactivate 99% of the cysts in 15 min at ph 7 and 5, respectively, at 25 C. These concentrations are attainable but well above the 0.5- to 1.0-mg/liter residuals typically employed in routine water treatment practice (26). Inactivation of 99% of cysts in less than 15 min with free chlorine requires more than 10 mg/liter at ph 9 and 25 C and more than 35 mg/liter at ph 7 and 5 C. Jarroll et al. (14) and Rice et al. (17) have described studies in inactivation of cysts of G. lamblia by chlorine. We products from the results of calculated approximate C. t' these two studies. At 250C, the C t' product was less than 15 mg min/liter for G. lamblia cysts compared with 25.5 to 44.8 mg min/liter for G. muris cysts. At 5 C, C. t' products for G. lamblia cysts were calculated to range from 90 to 170 mg min/liter compared with 449 to 1,012 mg min/liter for G. muris cysts. G. muris cysts are at least 1.5 times more resistant to chlorine than G. lamblia cysts at 25 C and are at least 2.5 times more resistant at 5 C. Thus, it can be concluded, as was found in an inactivation study with oone (27), that the more resistant G. muris is a good model for the human pathogen G. lamblia. G. muris cysts also appear to be more resistant to chlorine than N. gruberi cysts; average C- t' products of 7.72, 12.1,

5 1452 LEAHY ET AL. and 197 mg min/liter are reported for ph 5, 7, and 9, respectively, in experiments at 25 C (20). C. t' products obtained for G. muris cysts at the same temperature and similar contact times were in the ranges of 50.3 to 73.8 mg min/liter at ph 5, 25.5 to 44.8 mg min/liter at ph 7, and 174 to 224 mg min/liter at ph 9. Cysts of G. muris are therefore 6 to 10 times more resistant than cysts of N. gruberi at ph 5 and 2 to 4 times more resistant at ph 7. There were about equally resistant to chlorine at ph 9. C. t' products were approximated for cysts of Naegleria fowleri and two Acanthamoeba species from the data of De Jonckheere and van de Voorde (10). These cyst-forming protooans are the causative agents of amoebic meningoencephalitis. At 25 C, for chlorine residuals of 0.3 to 2.0 mg/liter, the C. t' products for N. fowleri cysts were in the range of 12 to 18 mg min/liter at ph 7.3 to 7.4 and 25 C. For 1.0 to 40.0 mg of chlorine per liter, the C. t' products for Acanthamoeba culbertsoni and the nonpathogenic Acanthamoeba sp. strain 4A were reported in the range of 2,500 to 7,500 mg min/liter and 960 to 5,000 mg min/liter, respectively. In contrast, the C. t' products for G. muris ranged from 25.5 to 44.8 mg min/liter for 2.80 to 7.12 mg of free chlorine per liter at ph 7. Accordingly, G. muris cysts are significantly more resistant than N. fowleri cysts, but are at least 20 times more sensitive to chlorine than Acanthamoeba sp. strain 4A and at least 50 times more sensitive than A. culbertsoni. G. muris cysts are approximately equal in resistance to chlorine as cysts of Entamoeba histolytica, the protooan responsible for amoebic dysentery. Stringer et al. (24) showed that 2 mg of chlorine per liter achieves 99% inactivation of Entamoeba cysts in about 10 min at ph 7 and 27 to 30 C. G. muris cysts required 16 min of contact time with 2.80 mg of chlorine per liter for the same level of inactivation at 25 C. Data from several studies are summaried in Table 2, providing a comparison to C. t' products for 99% inactivation of some protooan cysts, viruses, and bacteria by free chlorine. Some C. t' products were approximated from the reported data (Table 2, footnote a). These values are only roughly comparable because of differences in ph and temperature and, perhaps, general experimental technique. Protooan cysts are the most resistant to chlorine, however, and C. t' products for Escherichia coli are order of magnitude less resistant. Judging the microbiological safety of water solely on the basis of absence of coliforms is not a good practice. Results of this study indicate that chlorine inactivates cysts of G. muris, but only at concentrations well above these employed in routine water treatment practices. The relative ineffectiveness of chlorine against three cysts is exacerbated by conditions of high ph or low temperature. It can be concluded that G. lamblia may be modeled by G. muris, which is somewhat more resistant to free chlorine. These results also emphasie the need for very high dosecontact times for chlorination or for well-operated chemical treatment and subsequent filtration at conventional C t' levels for the effective removal or inactivation of protooan cysts in potable water supplies. ACKNOWLEDGMENTS This study was supported in part by the College of Engineering of The Ohio State University and by grant R from the Office of Exploratory Research, U. S. Environmental Protection Agency. LITERATURE CITED APPL. ENVIRON. MICROBIOL. 1. American Public Health Association Standard methods for the examination of water and wastewater, 15th ed. American Public Health Association, Washington, D.C. 2. Berg, G The virus haard in water supplies. J. N. Engl. Water Works Assoc. 78: Brett, R. W., and J. W. Ridgeway Experiences with chlorine dioxide in Southern Water Authority and Water Research Centre. J. Inst. Water Eng. Sci. 35: Center for Disease Control Waterborne giardiasis outbreaks-washington, New Hampshire. Morbid Mortal. Weekly Rep. 26: Center for Disease Control Giardiasis-Vail, Colorado. Morbid Mortal. Weekly Rep. 27: Centers for Disease Control Waterborne giardiasis- California, Colorado, Oregon, Pennsylvania. Morbid. Mortal. Weekly Rep. 29: Centers for Disease Control Outbreak of diarrheal illness associated with a natural disaster-utah. Morbid Mortal. Weekly Rep. 32: Cerda, J. J Giardiasis. Am. Family Physician 28: Craun, G. F Waterborne outbreaks. J. Water Pollut. Control Fed. 48: De Jonckheere, J., and H. Van de Voorde Differences in destruction of cysts of pathogenic and nonpathogenic Naegleria and Acanthamoeba by chlorine. Appl. Environ. Microbiol. 31: Engelbrecht, R. S., M. J. Weber, B. L. Salter, and C. A. Schmidt Comparative inactivation of viruses by chlorine. Appl. Environ. Microbiol. 40: Floyd, R., and D. G. Sharp Inactivation by chlorine of single poliovirus particles in water. Environ. Sci. Technol. 13: Jakubowski, W., T. H. Ericksen, and S. L. Chang Detection, identification, and enumeration of Giardia cysts in water supplies. In Proceedings of the American Water Works Association Water Technology Conference. American Water Works Association, Denver. 14. Jarroll, E. L., A. K. Bingham, and E. A. Meyer Effect of chlorine on Giardia lamblia cyst viability. Appl. Environ. Microbiol. 41: Jensen, H., K. Thomas, and D. G. Sharp Inactivation of coxsackievirus B3 and B5 in water by chlorine. Appl. Environ. Microbiol. 40: Meyer, E. A., and E. L. Jarroll Giardiasis. Am. J. Epidemiol. 111: Rice, E. W., J. C. Hoff, and F. W. Schaeffer III Inactivation of Giardia cysts by chlorine. Appl. Environ. Microbiol. 43: Ridgway, H. F., and B. H. Olson Chlorine resistance patterns of bacteria from two drinking water distribution systems. Appl. Environ. Microbiol. 44: Roberts-Thomson, I. C., D. P. Stevens, A. F. Mahmoud, and K. S. Warren Giardiasis in the mouse: an animal model. Gastroenterology 71: Rubin, A. J., J. P. Engel, and 0. J. Sproul Disinfection of amoebic cysts in water with free chlorine. J. Water Pollut. Control Fed. 55: Sawyer, C. N., and P. L. McCarty Chemistry for environmental engineering, 3rd ed. McGraw-Hill Book Co., New York. 22. Schaeffer, F. W., III, E. W. Rice, and J. C. Hoff In vitro excystation of Giardia muris cysts. Trans. R. Soc. Trop. Med. Hyg. 78: Sharp, D. G., and J. Leong Inactivation of poliovirus I (Brunhilde) single particles by chlorine in water. Appl. Environ. Microbiol. 40: Stringer, R. P., W. N. Cramer, and C. W. Kruse Comparison of bromine, chlorine and iodine as disinfectants for amoebic cysts. In J. D. Johnson (ed.), Disinfection-water and wastewater. Ann Arbor Science Publishers, Inc., Ann Arbor,

6 VOL. 53, 1987 Mich. 25. Taylor, G. R., and M. Butler A comparison of the virucidal properties of chlorine, chlorine dioxide, bromine chloride and iodine. J. Hyg. 89: Tchobanoglous, G., and E. D. Schroeder Water quality. Addison-Wesley Publishing Co., Reading, Mass. INACTIVATION OF GIARDIA CYSTS BY CHLORINE Wickramanayake, G. B., A. J. Rubin, and 0. J. Sproul Effect of oone and storage temperature on Giardia cysts. J. Am. Water Works Assoc. 77: Young, D. C., J. D. Johnson, and D. G. Sharp The complex reaction kinetics of ECHO-1 virus with chlorine in water. Proc. Soc. Exp. Biol. Med. 156: