Enumeration of Indicator Bacteria Exposed to Chlorine

Enumeration of Indicator Bacteria Exposed to Chlorine

Enumeration of Indicator Bacteria Exposed to Chlorine GORDONA. MCFETERSAND ANNE K. CAMPER Department of Microbiology, Montana State University, Bozema...

1MB Sizes 39 Downloads 120 Views

Enumeration of Indicator Bacteria Exposed to Chlorine GORDONA. MCFETERSAND ANNE K. CAMPER Department of Microbiology, Montana State University, Bozeman, Montana

I.

11.

111.

IV.

V.

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

191

I. Introduction The public health acceptability of water is evaluated by the presence of indicator bacteria. These microorganisms are widely employed to determine the potability of drinking water through the use of standardized test procedures to detect coliform bacteria (APHA, 1981). Numerical standards for these bacteria have been legislated (USEPA, 1976). At the same time, chlorine is added to most public drinking water supplies in the United States as an additional protective measure to disinfect the water prior to consumption. However, chlorine not only eliminates pathogens but also renders indicator bacteria incapable of forming colonies on currently accepted growth media. An underestimation of the actual number of indicator bacteria therefore results and a reduced or nonexistent potential public health threat is reported. The influence of chlorination on enumeration methodologies for waterborne sanitary indicator bacteria will be discussed. In addition, studies addressing the cellular mechanism of chlorine damage in coliform bacteria will be reviewed and improved procedures that maximize the recovery of these important microorganisms will be described.

II. Chlorination and the Enumeration of Waterborne Coliform Bacteria The application of chlorine in domestic water supplies has become one of the most widely used and effective public health measures in recorded history. Dramatic decreases in typhoid fever and other waterborne diseases 177 ADVANCES IN APPLIED MICROBIOLOGY, VOLUME 29 Copyright 0 1983 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-002629-5

178

GORDON A. MCFETERS A N D ANNE K . CAMPER

have accompanied the adoption of this procedure. This finding was first observed in Jersey City, New Jersey, just following the turn of the century and subsequently in Philadelphia, Pennsylvania, Wheeling, West Virginia, and many other communities across the United States. Documentation concerning this phenomenal success story is presented in the National Research Council’s “Summary Report: Drinking Water and Health” (United States National Academy of Sciences, 1977a), and in an early book by Hazen (1914). More recently (1938-1955), the national average of reported waterborne disease outbreaks continued to decline to a level of about 10 per year (United States National Academy of Sciences, 1977a). This laudable trend, however, has not continued and that number has tripled within the past 20 years. While a legitimate question remains whether this observation reflects a real trend or an artifact due to more agressive investigation and reporting, as suggested by Craun (1977), there is a persistent incidence of waterborne outbreaks within the United States. The actual number of waterborne outbreaks might, in fact, be underestimated due to incomplete reporting as well as nonspecific symptoms and the self-limiting nature of many such diseases. The majority of these outbreaks has been traced to deficiencies in water treatment systems and inadequate or interrupted chlorination. Because of these findings the conclusion may be drawn that the threat of disease has not been eliminated from our domestic water treatment and distribution systems despite the technological advances that have been made in engineering, microbiology, and public health. As a result, the need remains for continued progress in developing more accurate microbiological surveillance and analytical methodologies. These, in turn, will be useful in providing improved finished drinking water quality and safer wastewater discharges. At approximately the same time chlorination became widely used for the disinfection of municipal water sources, the bacterial indicator concept gained acceptance. The numerous advantages of using certain bacteria as an objective signal of a potential health hazard from fecal contamination soon became apparent. Although indicator bacteria are useful and extensively employed in connection with quality control of water and wastewater, this practice is not without some drawbacks. For instance, the indicator bacteria are isolated from the aquatic environment in which they are exposed to numerous chemical and physical factors that are stressful. The resulting injury may be significantly different for the indicator bacteria than for some of the pathogenic microorganisms (e.g., viruses) that represent the actual health hazard (Chambers, 1971; Ludovici et al., 1975; Berg et al., 1978). As a result, the indicators may not be enumerated in some instances where infectious viruses or other pathogens persist. Chlorinated waters represent a specific situation of this type because viruses in general have been characterized as being more resistant to chlorine and, therefore, as having greater

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

179

aquatic persistence than the indicator bacteria (Ludovici et al., 1975). Under these and other circumstances, where reliance is placed upon the enumeration of indicator bacteria as a criterion of water quality, maximized recovery of these microorganisms should be the primary goal in methodology development and testing. Problems and inadequacies associated with currently employed bacterial indicator systems such as coliforms (Dutka, 1973) have prompted efforts to consider alternative microorganisms or other concepts (Geldreich, 1977; White, 1978). However, despite the limitations of the current indicator microorganisms and enumeration methods, they remain the most reliable signal presently available to suggest the possible presence of waterborne fecal contamination and accompanying health hazards (United States National Academy of Sciences, 1977b). As a result, numerical coliform standards have been established for drinking water (USEPA, 1976) and secondary sewage effluent (USEPA, 1973) in the United States. As mentioned earlier, the presence of chlorine in water not only provides the desirable effects of reducing the total bacterial population and eliminating most of the pathogens, but it also alters the indicator bacteria (including total and fecal coliforms) in such a way that their enumeration is rendered less efficient (Camper and McFeters, 1978; Braswell and Hoadley, 1974). This is particularly a problem with the membrane filtration procedures commonly used to enumerate fecal coliform bacteria, because a selective growth medium (m-FC) is used at an elevated temperature (44.5”C). The selective conditions of media containing inhibitory compounds (as found in most coliform media) and the elevated temperature used in the case of fecal coliforms restrict the growth of chlorine-injured cells and a submaximal bacteria count is obtained (McFeters et al., 1982; Camper and McFeters, 1978). Such bacterial data are only partially representative of the actual viable indicator population. This underestimation is also seen with total coliform bacteria in water. For instance, a recent survey (unpublished data) revealed that 60-90% of the total coliform bacteria in the chlorine-treated drinking water from six small communities were injured. This results in enumeration efficiencies between 10 and 40%. Fecal coliform bacteria were also subject to cellular damage of greater than 94% when bacteria were exposed to chlorinated sewage for 5 minutes (Braswell and Hoadley, 1974). Under such circumstances, numerical standards and their enforcement become problematic. The time-honored multiple tube fermentation, or most probable number method (MPN), provides as much as 100 times greater coliform counts from chlorinated waters when compared with membrane filtration (MF) using selective media (Harold et al., 1956; McKee et al., 1958; Lin, 1973; Mowat, 1976; Moran and Witter, 1976; Schiemann et al., 1978). Both MF and MPN

180

GORDON A. MCFETERS AND ANNE K. CAMPER

procedures are listed in the current edition of “Standard Methods for the Examination of Water and Wastewater” (APHA, 1981), but because of the cumbersome nature of the MPN procedure and the imprecision of the information gained with this method, MF procedures are more extensively used for drinking water at the present time. The nonselective character of the initial MPN medium affords injured bacteria a suitable environment in which to recover and grow (McKee et al., 1958) if the noncoliform population is not excessive, i.e., < 500 ml. Because of that property, published reports that describe improved MF methods for the isolation of chlorineinjured bacteria (to be described in a subsequent section of this article) typically compare the data obtained on new media with that from the MPN procedure. It should also be pointed out that the MPN method is used more frequently to determine the coliform densities in sewage effluents and natural waters which contain elevated levels of suspended solids. Consequently, coliform populations determined with the MPN method might give a closer approximation of the actual number of indicator bacteria present under some conditions in which the number of noncoliforms is not excessive. Precision limitations of the MPN method, however, make comparisons of counts obtained with these two methods somewhat difficult. Considering the incomplete recovery of both total and fecal coliform bacteria from chlorinated drinking water and wastewater and the importance placed on the identification of these microorganisms, further discussion is warranted. Coliform bacteria are often used as a measure of disinfection efficiency in chlorinated drinking water and may be incompletely enumerated by currently accepted M F procedures (APHA, 1981). In a similar application, fecal coliform bacteria are considered a biological standard of disinfection efficiency in chlorinated wastewater (USEPA, 1973). The current microbiological enumerative procedures for the examination of both chlorinated drinking waters and wastewaters clearly represent trade-offs that are far from ideal because coliform indicator bacteria are either incompletely enumerated with M F methodology or are enumerated via the cumbersome and highly imprecise MPN procedure. It is not surprising, therefore, that dissatisfaction has been voiced concerning existing microbiological procedures. In that connection, other methods have been proposed to evaluate the efficiency of disinfection of wastewater (Tifft and Spiegel, 1976). The USEPA (1975) officially reconsidered the application of disinfection for wastewater, partially on the grounds that toxic and carcinogenic substances resulting from extensive chlorination (necessary to meet the standard) pose potential problems for aquatic ecosystems in receiving waters. This point was also made by Dugan (1978). A somewhat related situation exists for drinking water because it has been recommended that chlorine residual data may be substituted for 75% of the required coliform enumerations in some

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

181

cases due to incomplete bacteriological monitoring (National Interim Primary Drinking Water Regulations, 1976). It is interesting to note that this proposal has not met with much enthusiasm at the state level, where the actual decisions are made regarding such substitutions. In either case it should be emphasized that microbiological data describing the bacterial indicator populations in a given water or discharge represent useful biological information that is closely related to the potential health hazards that chlorination was implemented to control. This assertion is made in full cognizance of the inaccuracies and problems associated with the microbiological methodologies and a later section of this article will address those problems. Nonetheless, proposals to reduce the level of microbiological surveillance in water and wastewater would diminish this important source of information concerning the efficacy of the disinfection process because no alternative methods or organisms are currently accepted to assess this vital consideration. However, yeasts, acid-fast bacteria (Haas and Engelbrecht, 1980), and clostridia (Bonde, 1962) have been suggested because they exhibit greater resistance to chlorination than do coliforms.

111. Physiological Chlorine Injury in Waterborne Coliform Bacteria There have been a number of studies that have addressed the question of how chlorine injures or kills coliform bacteria. While most of these efforts were not carried out with any particular applied objective, this discussion will focus that information on the improvement of procedures for detecting coliform bacteria in chlorinated waters. In this way remedial measures may be considered for the specific physiological site of the chlorine-caused damage. In addition, the reversibility of chlorine injury must be documented and understood before improved recovery methodology may be considered a viable possibility. The potential reversibility of chlorine-induced damage in bacteria was suggested in 1935 by Mudge and Smith and in 1954 by Heinmets et al. when they reported using certain metabolic intermediates for the restoration of the viability of chlorine-inactivated Escherichia coli. This conclusion was challenged in papers from three different laboratories (Garvie, 1955; Chambers et al., 1957; Hurwitz et al., 1957), and was supported by others (McKee et al., 1958; Milbauer and Grossowicz, 1959). The importance of this consideration was recognized by microbiologists in the food industry, in which disinfectants such as chlorine and indicator organisms are also used. For example, Scheusner et al. (1971)published a paper in which they described the reversible injury of E . coli by chlorine and other sanitizers associated with foods.

182

GORDON A. MCFETERS AND ANNE K. CAMPER lo9

108

10'

. -E

v)

J J W

u

y.

0

10'

a W m I 3

z

lo!

10'

1

I

I

2

3

1

4

s

MINUTES

I

8

7

1

e

9

1

10

FIG.1. Injury of E . coli exposed by the laboratory chlorination procedure using a 0.5 mg/liter initial chlorine concentration at room temperature. Selective m-FC (0-0 incubation ) at 44.5% and nonselective overlay (0-0) at 35°C were used for enumeration.

Experiments were performed in our laboratories to examine the reversibility of chlorine-induced injury in E. coli, specifically under conditions that might be compared to chlorinated drinking water. In this study, a washed suspension of E. coli was resuspended in chlorine demand-free water at room temperature and chlorine (0.5 mglliter as sodium hypochlorite) was added. This chlorine concentration approximated the levels found in treated drinking water and wastewater. At timed intervals, the population was enumerated on m-FC medium incubated at 44.5"C and on a

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

183

rich nonselective medium at 35°C. The results, shown in Fig. 1, demonstrate that there was a rapid decrease in counts on the nonselective medium, representing the segment of the total population that was killed. Also, the differential between counts on the two media progressively increased with time of exposure in chlorine; that difference was interpreted as the part of the population that was injured, as discussed by Bissonnette et al. (1975). Similar results have also been found with total coliforms (Camper and McFeters, 1978). These results indicate that coliform suspensions exposed to chlorine are rapidly killed but a sizable fraction of the survivors are injured in such a way that they fail to grow under the selective conditions commonly used to detect these organisms in water samples. Also, the injured population may be as large as !30% of the total surviving population. In the companion experiment, E. coli suspensions that had been exposed to chlorine (0.5 mg/liter) for 2.5 hours were added to a rich nonselective medium at 35°C. At timed intervals, aliquots were removed and the bacteria enumerated on the two media, as described in the first experiment. Control cells that had not been exposed to chlorine were also inoculated into the nonselective medium and enumerated as were the chlorinated cells. The data in Fig. 2 indicate that the control bacteria exhibited a short lag period and then began logarithmic growth showing equal counts on the two media. The chlorinated cells, however, demonstrated a different response. A prolonged period was seen during which the counts on the two media were different. After 4 hours the cells commenced logarithmic growth at a rate that was equivalent to the control suspension and the counts on the two media became identical. These findings demonstrate that chlorine-injured cells are capable of recovery in a rich nonselective medium, as evidenced by their persistence during the protracted lag phase. Following this repair process, the injured bacteria were equally capable of growth on selective and nonselective media at rates comparable to those of uninjured cells. The implications of these findings relative to the present discussion are that the majority of the cell population is incapable of growth on m-FC medium at 44.5”C, but if these cells are provided with a nonselective medium and temperature in which to recover for 4 hours, they regain the ability to grow and form colonies on the selective medium at the elevated temperature. These results also suggest that improved methods may lead to greater recovery of chlorine-injured cells from treated water and wastewater. Another implication of these experiments relates to the “regrowth” phenomenon commonly seen as an increase in numbers of coliform bacteria some distance downstream from a treatment plant outfall. McKee et al. (1958) and others (Shuval et al., 1973) suggested that the rejuvenation of chlorine-injured cells from treatment plant effluents might be responsible

184

GORDON A. MCFETERS A N D ANNE K. CAMPER

/

---d FI

1

FIG. 2. Repair in overlay broth of E . coli cells that were exposed to chlorine (0.5 mg/liter, ) initial concentration) for 8 minutes. Control cells (0-0) and chlorinated cells (0-0were enumerated over a 6-hour growth period using overlay (-) and m-FC ( - - - ) media and the membrane filter technique.

for the observation. Schillinger and Stuart (1976), Camper (unpublished data) in our laboratories, and others (Kinney et al., 1978) also reached this conclusion. The physiological mechanism of chlorine damage in E . coli was reported in the mid-1940s as an irreversible oxidation of sulfhydryl-containing enzymes in general, and aldolase in particular (Green and Stumpf, 1946; Knox et aZ., 1948). Ingols et al. (1953) supported this hypothesis indirectly by demonstrating that chlorine reacted with cysteine. Skidal'skaya (1969) and

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

185

Venkobachar et aZ. (1975) performed experiments investigating the effect of chlorine on bacterial dehydrogenases. In all of these studies the implication that chlorine injury was due to sulfhydryl group oxidation was drawn from in vitro chlorination procedures and in vitro decreases in dehydrogenase activity that correlated with reduced cellular viability. Studies were carried out in our laboratory to further examine the effect of chlorination on the level of aldolase remaining in cell-free extracts prepared from cell suspensions that had been chlorinated prior to cell disruption (Camper and McFeters, 1978). When these enzyme activities were compared with comparable preparations from nonchlorinated cells, there was no statistical difference. Our findings, therefore, indicated that when intact E . coli cells were exposed to 0.5 mg/liter chlorine for 8 minutes, intracellular aldolase activity was not altered. Additional experiments conducted in our laboratories to examine the physiological effects of chlorine on E . coli (Camper and McFeters, 1978) were also directed at the oxygen uptake and intracellular ATP levels following chlorination. The response of these cellular parameters to chlorination was similar to that seen for viability; an initial rapid decline was followed by a gradual reduction. Also, the addition of reducing agents or the presence of more complex media provided no remedial effect. This would suggest that a simple chemical reduction of chlorine-oxidized sites is not the primary mode of recovery. The next approach in our study examined the possibility that chlorine acted on or near the cell surface. Bacterial cell envelopes contain many potential chlorine-reactive sites that are exposed to the extracellular environment although electron microscopic studies of cells treated with chlorine demonstrated no morphological alteration in their envelope structure (Bringman, 1953). Our experiments, employing “Clabeled glucose and amino acids, encompassed two related concepts: the ability of chlorinated cells to take up exogenous substrates and the metabolic turnover of incorporated compounds in treated cells. If cell envelope damage was involved in chlorine injury, the uptake of labeled substrates would likely be impaired. The first group of experiments examined the uptake of 14C-labeled glucose following chlorination. A striking difference between uptake in control and chlorinated cells was observed (Table I). This indicates that chlorine-treated cells had compromised carbohydrate transport systems and were incapable of moving glucose across the cell envelope. To test the possibility that chlorination decreased the intracellular turnover of materials derived from glucose, cells were labeled with [14C]glucosefor 1 hour prior to chlorination. The results of these experiments revealed no difference in cellular radioactivity between control and chlorinated suspensions during the 120 minutes following chlorine exposure, suggesting that net carbon metabolism in chlo-

186 COMPARISON OF

GORDON A. MCFETERS A N D ANNE K. CAMPER

R\DIOACTIVITY

TABLE I UPTAKE IN CONTROL AND %MINUTE CHLORINE-TREATED CELLSI N TSY AT 3!5"Ca Control

Minutes

0 5 10 15 20 25 30 Background

External standard

6.21 6.14 6.22 6.10 6.22 6.08

6.30 6.21

Chlorinated

CPm

200.93 197.34 190.47 196.71 197.36 163.39 146.59 13.83

External standard

CPm

6.21 6.I1 6.17 6.12 6.10 5.99 6.02

13.95 13.95 13.95 14.12 14.89 14.38 14.37

C e l l s were labeled with [14C]glucosefor 10 minutes following chlorination or no treatment (control).

rinated cells continued at about the same rate as before exposure to the disinfectant. It was anticipated that the turnover of intracellular reserve materials would be accelerated to facilitate repair processes during the chlorine-imposed discontinuation of carbohydrate uptake. However, intracellular storage materials utilized for this repair function had apparently not been appreciably labeled in our experiments. Further investigation was needed to determine if the effect of chlorine on cellular transport was limited to a carbohydrate system or if it was more general. Labeled algal protein hydrolysate was used because amino acids are also transferred across the bacterial envelope via active transport systems. A repeat of the postchlorination uptake experiment with algal protein hydrolysate (largely amino acids) yielded essentially the same results as found with glucose, indicating the nonspecific nature of chlorine damage to bacterial transport processes. Other workers have proposed a variety of chlorine-induced abnormalities in bacterial physiology. One such report suggests that chlorine-caused death is due to unbalanced metabolism after disruption of part of the cellular enzyme system, and that recovery may occur when the damaged enzymes are replaced (Wyss, 1961). Protein synthesis was also implicated when it was reported that chlorine damage is a multihit process resembling the disruption of that cellular mechanism (Benarde et al., 1967). Chlorine has also been reported to change the structure of amino acids (Pereira et al., 1973). A more comprehensive study reported by investigators in India (Venkobachar et al., 1977) indicated that oxygen uptake and oxidative phosphorylation were in-

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

'

187

hibited by chlorine and that chlorination caused the leakage of some macromolecules in a B strain of E . coli. Because the permeation of exogenous substrates into bacteria is of major importance in cellular metabolism and because this process has been shown to be severely inhibited by chlorine, the ability to uptake nutrients is proposed as the primary physiological lesion in chlorinated E. coli cells. Resultant metabolic consequences would likely reflect unbalanced metabolism as well as reduced levels of protein synthesis and energy-related parameters like oxygen uptake, ATP concentration, and oxidative phosphorylation. On that basis, we proposed (Camper and McFeters, 1978) the cellular injury resulting from exposure ofE. coli to chlorine centers around the compromise of substrate uptake and the degree of damage, ranging from reversible injury to death, is dependent upon factors such as chlorine concentration and time of exposure. That conclusion is supported by reports in the literature indicating that chlorine inhibition is a multihit process (Benarde et al., 1967; Fair et al., 1958) that inhibits uptake and related physiological processes (Kulikovsky et al., 1975; Haas, 1980). In addition to these physiological consequences of exposing bacteria to chlorine in water, some genetic effects have been reported. Chlorine has been shown to react with and modify purine and pyrimidine bases (Patton et al., 1972; Hoyano et al., 1973). More recent studies have also pointed to the bacterial mutagenic potential of chloramine (Shih and Lederberg, 1976a,b). These effects suggest many far-reaching microbiological implications of chlorination in aquatic systems.

IV. Improved Enumeration of Bacteria in Chlorinated Waters As described earlier, problems associated with the incomplete recovery of coliform indicator organisms from chlorine-containingwater and wastewater are well recognized. This inaccuracy represents a significant underestimation of waterborne bacterial numbers because the MF procedure for the enumeration of coliform bacteria may yield between 10- and 100-fold fewer bacteria than the MPN method. This point was made in a report by Braswell and Hoadley (1974) in which they indicated the imperative need for more reliable methods of bacterial enumeration in chlorinated waters. Maxey (1973) also pointed out other similar situations. The possibility of improved MF methods is suggested by the reversible nature of bacterial chlorine injury (Camper and McFeters, 1978; McKee et al., 1958; Milbauer and Grossowicz, 1959; Maxey, 1973). The general conceptual framework of such an advance was presented in a review article by Harris (1973), outlining the approach for recovering sublethally injured

188

GORDON A. MCFETERS A N D A N N E K. CAMPER

cells. He proposed that such bacteria must be maintained initially in a nonselective chemical and physical environment to allow “revival.” Following that resuscitation period, the cells regain the ability to grow under more rigorous conditions, such as on the selective media and at the temperature used in the M F procedure for the enumeration of fecal coliform bacteria (Geldreich et al., 1965). A. FECAL COLIFORMBACTERIA Prompted by the inherent limitations of the MPN method for the enumeration of fecal coliform bacteria and the incomplete recovery of these organisms in chlorinated waters by MF methods, workers considered ways the basic m-FC medium and procedure of Geldreich et al. (1965) could be improved. The objective of the task was to modify the selective m-FC method to include an initial nonselective step and by so doing produce a protocol encompassing the desirable characteristics of membrane filtration with improved recovery of injured bacteria. A two-step enrichment procedure involving a preincubation of the filter with the bacteria on a nonselective medium has been listed for some time in “Standard Methods for the Examination of Water and Wastewater”; however, it has not been extensively used. Another hybrid procedure was suggested by Greene et al. (1974) in the form of a 2- to 6-hour preincubation on dilute m-FC medium at 25°C followed by transfer to conventional medium at 44.5”C for the remaining 18 to 22 hours. These modifications resulted in improved recovery of fecal coliforms from chlorinated waters. Because of these findings and the previous suggestion (McKee et al., 1958) that the damage in chlorine-injured cells may be reversed, Rose et al. (1975) developed a modification of the previous method. This procedure included a two-layer medium, prepared immediately prior to use, in which the upper layer was nonselective and the bottom layer was selective. In this way the injured bacteria on the filter were initially exposed to a nonselective nutrient environment and incubated at 35°C for 2 hours after which time the temperature was raised to 44.5”C for the remaining 22 to 24 hours. The selective agents diffused upward through the top layer within 3 to 4 hours and therefore prevented the overgrowth of noncoliform bacteria on the filter. Comparative results indicated that this procedure was successful in yielding greater bacterial numbers from chlorinated waters with a fecal coliform verification rate of greater than 90%. However, the original paper did not present sufficient data from chlorinecontaining waters and did not compare the new method with the MPN method. Further tests indicated that the MPN method still produced 10-fold more bacteria than the new method (Green et al., 1977) and therefore indicated a need for further improvement. An alternative approach has been

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

189

suggested by the use of a timed-release capsule containing the selective agents used in MPN methods (Lanz and Hartman, 1976). Dissatisfaction with M F materials and procedures began to reach critical mass shortly after Rose et aZ. (1975) described their method. That unrest led the USEPA to make the suggestion that MF methods should be reconsidered as acceptable water quality assessment tools under some circumstances. These developments prompted individuals within the American Society for Testing and Materials and the USEPA to organize a meeting entitled “Symposium on the Recovery of Indicator Organisms Employing Membrane Filters” that was held in January, 1975 (USEPA, 1977). One result of that meeting was the involvement of three separate laboratories in the development of new recovery methods and media. Each of these groups has developed a new methodology for the improved recovery of fecal coliforms from chlorine-treated water and wastewater (Green et d.,1977; Lin, 1976; Stuart et al., 1977). All of these methods have been evaluated and yield results that are comparable with MPN data for fecal coliforms from chlorinated water. The first was developed by Lin (1976) and included two separate media and two temperatures of incubation. The second, by Green et a2. (1977), involved a modified temperature incubation schedule of 5 hours at 35°C followed by 18 hours at 44.5”C. The simplicity of this modified procedure provides a clear practical advantage, particularly since Millipore Corporation (Bedford, Mass.) has marketed a temperature-programmed incubator specifically for that purpose. The third method, proposed by Stuart et al. (1977), embodied two temperatures and a two-layer medium, the overlay containing a rich mixture of metabolic intermediates. As with the others, this procedure yields results from chlorinated waters that compare favorably with MPN data and an acceptably high fecal coliform confirmation rate. B. TOTALCOLIFORMBACTERIA Currently accepted methods (APHA, 1981) for the enumeration of total coliform bacteria have several shortcomings, including low recoveries of injured cells (Dutka, 1973; McFeters et al., 1982). This observation prompted us to develop a new medium for the improved recovery of total coliform bacteria from chlorinated drinking water (LeChevallier et al., 1982). This medium, called m-T7, has been evaluated with contaminated drinking water samples and yields nearly three times more coliforms than the currently accepted m-Endo medium (APHA, 1981)with acceptable confirmation and levels of false negatives and false positives. These results are promising, but additional testing will be needed in other geographical areas to fully evaluate this new medium.

190

CORDON A. MCFETERS A N D A N N E K. CAMPER

C. FECAL STREPTOCOCCUS BACTERIA Stress has been implicated as a factor that can reduce the recovery efficiency of these bacteria from waters and wastewaters, particularly those containing chlorine (Rose and Litsky, 1965; Lin, 1974). A comparative study of two-step preenrichment techniques that has been proposed (Lin, 1974) indicates that the use of bile broth medium yields fecal streptococcus recoveries that are comparable with MPN results from the same samples. This technique involves preincubation of membrane filters containing bacteria on an enrichment medium for 2 hours at 35°C followed by plating on M-Enterococcus agar for 48 f 2 hours at 35°C. It should be noted that the above methodological developments were made without the benefit of the recent and more definitive information on the mechanism of bacterial chlorine injury. It is hoped that the development of more effective water quality assessment methods will now be possible.

V. Summary Stress resulting from a variety of chemical and physical environments has been recognized in indicator bacteria. A review by Busta (1976) summarizes the extensive work that has been carried out to describe indicator microorganisms sublethally impaired due to a variety of causes associated with foods. Workers in the area of water microbiology are also gaining an appreciation of the importance of these stressed cells in the assessment of water quality using bacterial indicators. Chemical agents, including chlorine, that are employed in water disinfection processes are important causes of bacterial stress injury. As a result, a significant portion of the total population of indicator bacteria in water might not be enumerated (using the selective procedures that are currently employed) and inaccurate water quality determinations could result. Alternative water disinfection agents that are being suggested, such as ozone, chlorine dioxide, and ultraviolet irradiation, will also probably lead to the same result. In addition, heat from thermal pollution and interactions with other microorganisms or chemicals (including disinfectants and metals) also exert stress that could further debilitate indicator bacteria in various waters and effluents. A need for improved enumeration procedures has accompanied the recognition of injured indicator bacteria in chlorinated waters and wastewaters. This movement has also stimulated interest in the underlying mechanism of cellular damage that is responsible for the submaximal recovery of coliforms from disinfected waters. Various groups have reported that a number of biochemical, genetic, and physiological processes are impaired by chlorine exposure under differing conditions. Evidence from our laboratory and else-

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

191

where implicates functions associated with the cell envelope, i.e., the uptake of extracellular organic substrates, as the primary cellular target of chlorine under conditions that are similar to those in the field. Additional data from our group indicate that sublethal damage from chlorine can be reversed under suitable nonselective conditions. Recent efforts have led to the development of new methods to enumerate injured fecal streptococcus, total and fecal coliform bacteria from chlorinated waters and wastewater. These procedures each yield data that are comparable with that obtained using the more cumbersome MPN method. As a result, the best characteristics of both methods may now be found in three relatively simple M F procedures. Some of these advances have been described in a new section (#921) of the fifteenth edition of “Standard Methods for the Examination of Water and Wastewater’’ entitled “Stressed Organisms” (APHA, 1981). However, it is anticipated that new and better water quality assessment methodologies will emerge from the growing literature concerning the physiological and biochemical behavior of indicator microorganisms in water and wastewater. These important advances will not only aid in guaranteeing the efficiency of aquatic disinfection processes but will provide more precise bacteriological information required to carry out meaningful epidemiological studies of waterborne disease outbreaks. Armed with these vital data, environmental scientists, public health officials, and the operators of individual wastewater treatment plants and domestic potable water systems, will be able to provide even greater protection for the public. REFERENCES American Public Health Association. (1981). “Standard Methods for the Examination of Water and Wastewater,” 15th ed. Am. Public Health Assoc., New York. Benarde, M . , Snow, W. B., Olivieri, V. P., and Davidson, B. (1967). Appl. Microbiol. 15, 257-265. Berg, G . , Dahling, D . R., Brown, G . A., and Berman, D.(1978).A p p l . Enoiron. Microbiol. 36, 880-884. Bissonnette, G . K . , Jezeski, J. J., McFeters, G . A., and Stuart, D. G.(1975). A p p l . Microbiol. 29, 186-194. Bonde, G . K. (1962). “Bacterial Indicators of Water Pollution.” Teknisk Forlag, Copenhagen. Brasswell, J. R., and Hoadley, A. W. (1974). A p p l . Microbiol. 28, 328-329. Bringman, G. (1953). Z. Hyg. Znfektionskr. 138, 155-156. Busta, F. F. (1976). J . Milk Food Technol. 39, 138-145. Camper, A. K.,and McFeters, G. A. (1978). A p p l . Enoiron. Microbiol. 37, 633-641. Chambers, C.W. (1971). J . Water Pollut. Control Fed. 43, 228-241. Chambers, C.W., Tabak, H. H., and Kabler, P. W. (1957). J . Bacteriol. 73, 77-84. Cram, G . F. (1977). J . Water Pollut. Control Fed. 49, 1268-1279. Dugan, P. R. (1978). A m . SOC. Microbiol. News 44, 97-102.

192

GORDON A. MCFETERS A N D ANNE K. CAMPER

Dutka, B. J. (1973).J . Enoiron. Health 36, 39-46. Fair, G. M., Morris, J. C., Chang, S. L., Weil, I., and Burden, R. P. (1958).J . Am. Water Works Assoc. 40, 1051-1061. Garvie, E. I. (1955).J . Bacteriol. 67, 393-398. Geldreich, E. E. (1977).J. Water Pollut. Control Fed. 49, 1222-1244. Geldreich, E. E., Clark, H. F., Huff, C. B., and Best, L. C. (1965).J . Am.Water Works Assoc. 57, 208-213. Green, B. L., Clausen, E. M., and Litsky, W. (1977). Appl. Enuiron. Microbiol. 33, 1259-1264. Green, D. E., and Stumpf, P. K. (1946).J . Am. Water Works Assoc. 38, 1301-1305. Greene, R. A., Bordner, R. H., and Scarpino, P. V. (1974). Abstr. Annu. Meet. Am. SOC. Microbiol., p. 34. Haas, C. N. (1980).J . Water Pollut. Control Fed. 52, 1976-1989. Haas, C. N., and Engelbrecht, R. A. (1980). Water Res. 14, 1749-1758. Harold, A. T., Woodward, R. L., and Kabler, P. W. (1956).J . Am. Water Works Assoc. 48, 1391-1402. Harris, N. D. (1973).J . Appl. Bacteriol. 26, 387-397. Hazen, A. (1914). “Clean Water and How To Get It,” 2nd ed. Wiley, New York. Heinmets, F., Taylor, W. W., and Lehman, J. J. (1954).J . Bacteriol. 67, 5-12. Hoyano, Y., Bacon, V., Summons, R. E., Periera, W. E., Halpern, B., and Duffield, A. M. (1973). Biochem. Biophys. Res. Commun. 53, 1195-2001. Hurwitz, C., Rosano, C. L., and Blattberg, B. (1957).J . Bacteriol. 73, 743-746. Ingols, R. S., Wyckoff, H. A., Kethley, T. W., Hodgden, H. W., Fincher, E. L., Hildebrand, J. C., and Mandel, J. E. (1953). Ind. Eng. Chem. 45, 996-1000. Kinney, E.C., Drummond, D. W., and Hanes, N. B. (1978).J . Water Pollut. Control Fed. 50, 2307-32 12. Knox, W. E., Stumpf, P. K., Green, D. E., and Auerbach, Y. H. (1948). J . Bacteriol. 55, 451-458. Kulikovsky, A., Prakaratz, H. S . , and Sadoff, H. L. (1975).J . Appl. Bacteriol. 38, 39-46. Lanz, W. W., and Hartman, P. A. (1976). Appl. Enoiron. Microbiol. 32, 716-722. LeChevallier, M. W., Cameron, S. C., and McFeters, G . A. (1982).Appl. Enoiron. Microbiol. 43, 97-103. Lin, S. D. (1973).J . Water Pollut. Control Fed. 45, 498-506. Lin, S. D. (1974).J . Enoiron. Eng. Diu. Proc. Am. SOC. Ciu. Eng. LOO, 253-257. Lin, S. D. (1976). Appl. Enoiron. Microbiol. 32, 547-552. Ludovici, P. P., Phillips, R. A., and Jeter, W. S. (1975).In “Disinfection, Water and Wastewater” (J. D. Johnson, ed.). Ann Arbor Science, Ann Arbor, Michigan. McFeters, G. A., Cameron, S. C., and LeChevallier, M. W. (1982). Appl. Microbiol. 43, 97-103. McKee, J. E., McLaughlin, R. T., and Lesgourgues, P. (1958). Sewage Ind. Wastes 30, 245-252. Maxey, R. B. (1973).J. Milk Food Technol. 36, 414-421. Milbauer, R., and Grossowicz, N. (1959). Appl. Microbiol. 7, 67-70. Moran, J. W., and Witter, L. D. (1976). Water Sewage Works 123, 66-67. Mowat, A. (1976).J . Water Pollut. Control Fed. 48, 724-728. Mudge, C. S., and Smith, F. R. (1935). Am. J . Public Health 25, 442-447. National Interim Primary Drinking Water Regulations-Part 141. (1976).J . Am. Water Works Assoc. 68, 57. Patton, W., Bacon, V., Duffield, A. M., Halpern, B., Hoyano, Y., Pereira, W., and Lederbert, J. (1972). Biochem. Biophys. Res. Commun. 48, 880-884.

ENUMERATION OF BACTERIA EXPOSED TO CHLORINE

193

Pereira, W. E., Hoyano, Y., Summons, R. E., Bacon, V. A., and DuEield, A. M. (1973). Biochim. Biophys. Acta. 313, 170. Rose, R. E., and Litsky, W. (1965). Appl. Microbiol. 13, 106. Rose, R. E., Geldreich, E. E., and Litsky, W. (1975). Appl. Microbiol. 29, 532-536. Scheusner, D. L., Busta, F. F.,and Speck, M. L. (1971). Appl. Microbiol. 21, 4-45. Schiemann, D. A., Brodsky, M. H., and Ciehin, B. W. (1978). J. Water Pollut. Control Fed. 50, 158-162. Schillinger, J. E., and Stuart, D. G. (1976). Abstr. Annu. Meet. Am. Soc. Microbiol., p. 172. Shih, K. L., and Lederherg, J. (1976a).J . Bacteriol. 125, 934-945. Shih, K. L., and Lederberg, J. (1976h). Science 192, 1141-1143. Shuval, H. I., Cohen, J., and Kolodney, R. (1973). Water Res. 7 , 537-546. Skidal’skaya, A. M. (1969). Gg. Sanit. 34, 11-17. Stuart, D. G . , McFeters, G . A , , and Schillinger, J. E. (1977). Appl. Enoiron. Microbiol. 34, 42-46. Tifft, E. C., and Spiegel, S. J. (1976). Enoiron. Sci. Technol. 10, 1268-1272. United States Environmental Protection Agency (USEPA) (1973). Fed. Register 38, 10642. United States Environmental Protection Agency (USEPA). (1975). Fed. Register 40, 34522. United States Environmental Protection Agency (USEPA). (1976). National Interim Primary Drinking Water Regulations. EPA-570/9-76-003. Washington, D.C. United States Environmental Protection Agency (USEPA). (1977). “Symposium on the Recovery of Indicator Organisms Employing Membrane Filters” (R. H. Bordner, C. F. Frith, and J. A. Winters, eds.). EPA-609/9-77-024. Otfice of Research and Development, Environmental Monitoring and Support Laboratory, Cincinnati, OH, 45268. United States National Academy of Sciences. (1977a). “DrinkingWater and Health,” a report of the Safe Drinking Water Committee, Advisory Center on Toxicology, Assembly of Life Sciences, National Research Council, Washington, D.C. United States National Academy of Sciences. (1977h). “Summary Report: Drinking Water and Health,” a Report of the Safe Drinking Water Committee, Advisory Center on Toxicology, Assembly of Life Sciences, National Research Council, Washington, D.C. Venkobachar, C., Iyengar, L., and Rao, A. V. S. P. (1975). Water Res. 9, 119-124. Venkohachar, C., Iyengar, L., and Rao, A. V. S. P. (1977). Water Res. 11, 727-729. White, G . C. (1978).J . Am. Water Works Assoc. 70, 219-222. Wyss, 0. (1961). In “Public Health Hazards of Microbial Pollution of Water.” Proc. Rudolfs Res. Conf., Rutgers University, New Brunswick, New Jersey.