Survival of Bacterial Indicator Species and Bacteriophages after Thermal Treatment of Sludge and Sewage

Abstract

The inactivation of naturally occurring bacterial indicators and bacteriophages by thermal treatment of a dewatered sludge and raw sewage was studied. The sludge was heated at 80°C, and the sewage was heated at 60°C. In both cases phages were significantly more resistant to thermal inactivation than bacterial indicators, with the exception of spores of sulfite-reducing clostridia. Somatic coliphages and phages infecting Bacteroides fragilis were significantly more resistant than F-specific RNA phages. Similar trends were observed in sludge and sewage. The effects of thermal treatment on various phages belonging to the three groups mentioned above and on various enteroviruses added to sewage were also studied. The results revealed that the variability in the resistance of phages agreed with the data obtained with the naturally occurring populations and that the phages that were studied were more resistant to heat treatment than the enteroviruses that were studied. The phages survived significantly better than Salmonella choleraesuis, and the extents of inactivation indicated that naturally occurring bacteriophages can be used to monitor the inactivation of Escherichia coli and Salmonella.

Sewage depuration results in the production of sludge. Among other potentially noxious substances, sludge produced by wastewater treatment plants that treat fecal material contains pathogenic microorganisms. The most important pathogenic microorganisms are those transmitted by the fecal-oral route and include bacteria, viruses, and parasites. Fecal microorganisms that are pathogenic or nonpathogenic and are present in sewage are transferred to sludge, where their concentration is higher than it is in wastewater (7, 18, 28, 31, 38). Moreover, many microorganisms survive better when they are associated with solids than when they are suspended in water (20, 27, 30, 36), and at some stage of treatment of sludge, the growth of some pathogenic bacteria (for example, Salmonella) may be supported (13). Therefore, both treated sludge and raw sludge are sources of pathogenic microorganisms.

Legislation in industrialized countries requires acceptable procedures for the disposal and utilization of sludges (12, 35). One of these procedures, land application, offers many advantages, some of which are the disposal of cumbersome material, supply of valuable nutrients to agricultural crops, and improvement of soil properties. However, there is public concern about the agricultural use of sludge. Agricultural sludge reuse is acceptable and accepted only if the sanitary quality of the sludge is sufficiently guaranteed and the public concern can be limited. States or multinational organizations have implemented or are implementing regulations which are based on our present knowledge (12, 34, 35). Regarding pathogens, the regulations for use of sludges for agricultural purposes are based on three principles: (i) a requirement for treatment, including thermal treatment, to reduce the amount of pathogens; (ii) validation of the treatment; and (iii) assurance of the microbiological quality of the sludges.

Ideally, validation of a treatment and assurance of the microbiological quality of sludges should be done with pathogens, but this is not feasible at this time. Although powerful molecular methods for the detection of pathogens have been developed, at the present stage of development these methods cannot be used to measure the performance of treatments since they are unable to discern viable microorganisms from nonviable microorganisms (1). Therefore, surrogate indicators (e.g., fecal coliform bacteria or Escherichia coli) should be used for routine evaluation of treatment plant performance and sludge quality. Because of the different extents of inactivation exhibited by various microorganisms after treatments (7, 11, 16, 30, 38), it seems desirable to consider adding some indicator or indicators to E. coli or fecal coliform bacteria to guarantee lower health risk.

Somatic coliphages, F-specific RNA phages, and bacteriophages that infect Bacteroides fragilis have been proposed as potential indicators of water quality and/or virus content (14), and spores of sulfite-reducing bacteria have been proposed as indicators of parasitic protozoan content (25).

Salmonella choleraesuis subsp. choleraesuis serovar Senftenberg has also been suggested as a suitable test organism for validation of the processes based on thermal treatment (26, 31). However, in a full-scale treatment plant it may not be practical to inoculate the sludge, although some authors have proposed various forms of water-permeable bags that could be used to expose the test organisms to the treatment conditions (26).

We present here the results of thermal treatment experiments performed in the laboratory with sludge or raw sewage to compare the survival of different indicators, including enteroviruses. Naturally occurring indicators, as well as microorganisms previously grown in the laboratory that were added to sludge and raw sewage, were studied. Special attention was given to phages, which in spite of their potential as surrogate indicators of water quality (14) have been hardly studied in sludges.

MATERIALS AND METHODS

Bacteria, bacteriophages, and viruses.

S. choleraesuis subsp. choleraesuis serovar Senftenberg strain CECT 4384 was used in experiments in which thermal treatment of amended samples was examined.

Bacteriophage φX174 (ATCC 13706-B1), which belongs to the family Microviridae, was used as the reference bacteriophage for somatic coliphages (5). Bacteriophage MS2 (ATCC 15597-B1), which belongs to the family Leviviridae, was used as the reference bacteriophage for F-specific RNA bacteriophages (3), and bacteriophage B40-8 (ATCC 51477-B1), which belongs to the family Siphoviridae, was used as the reference bacteriophage for phages infecting B. fragilis HSP40 (33). Other somatic coliphages included in this study (MY2, SR51, SC12, and SS13) were isolated from environmental samples, characterized by electron microscopy, and propagated as previously described (10). Briefly, phage MY2 belongs to the family Myoviridae, and phages SR51, SC12, and SS13 are members of the Siphoviridae with straight, flexible, and curled tails, respectively. Phage suspensions were diluted in phosphate-buffered saline.

Attenuated poliovirus type 1 strain Lsc-2ab and five environmental isolates, three of which were identified as coxsackievirus B4, coxsackievirus B5, and echovirus 6 by nested PCR followed by restriction fragment length polymorphism (17) and two of which were enteroviruses (EV 1 and EV 2) that were not fully identified by the technique mentioned above, were used in this study. Viruses were propagated in the BGM continuous cell line. The cells were grown in Eagle's minimum essential medium (MEM autopow; ICN Biomedicals Inc., Aurora, Ohio) containing 5% fetal bovine serum, 2 mM l-glutamine, 26.8 mM NaHCO₃, 100 U of penicillin ml⁻¹, and 100 μg of streptomycin ml⁻¹.

Samples.

Sludge and raw sewage samples were obtained from the wastewater treatment plant of Gavà-Viladecans, which is located 10 km south of Barcelona, Spain. This plant depurates domestic and industrial wastewater from a population of 400,000 inhabitants. The treatment includes primary sedimentation and aerobic activated sludge digestion. A mixture of primary sludge (two-thirds) and activated sludge (one-third) undergoes anaerobic digestion at 35°C for about 25 days; then it is treated with a solution of synthetic organic polyelectrolyte flocculant prior to mechanical dehydration by centrifugation until the moisture content is 75%. The incoming raw sewage and the dewatered sludge were used in this study.

Raw sewage and sludge samples were collected in sterile containers, transported to the laboratory within 2 h after collection, and stored at 4°C (maximum storage time, overnight) until they were tested for bacteria, phages, and viruses.

Extraction of microorganisms from sludge.

Ten grams of sludge was suspended in 100 ml of 0.25× Ringer's solution and stirred with a magnetic bar for 15 min at room temperature. After centrifugation at 1,500 × g for 15 min at 4°C, the supernatant was used as a liquid sample for bacterial enumeration.

Ten grams of sludge was suspended in 100 ml of a 10% (wt/vol) beef extract solution and stirred with a magnetic bar for 15 min at room temperature. After centrifugation at 1,500 × g for 15 min at 4°C, the supernatant was filtered through a 0.22-μm-pore-size low-protein-binding polyvinylidene fluoride membrane (Millex-GV; Millipore) to remove bacteria before phage enumeration (32).

Microbial enumeration.

Bacteria were counted by using membrane filter assays for fecal coliforms and fecal streptococci according to the American Public Health Association standard methods (2). E. coli contents were determined after filter assays on Fluorocult agar (Difco). Spores of sulfite-reducing clostridia (SRC) were counted as described by Bufton (9). S. choleraesuis contents were determined by membrane filter assays by using MacConkey agar (Difco) for detection of yellow colonies.

Bacteriophages were counted by the double-agar-layer technique by using the ISO 10705-2 standard (5) for enumeration of somatic coliphages, the ISO 10705-1 standard (3) for enumeration of F-specific RNA bacteriophages, and the ISO 10705-4 standard (6) for enumeration of bacteriophages infecting B. fragilis.

Enteroviruses were enumerated after adsorption to cellulose nitrate membrane filters as previously described (24).

Amendment of samples.

Ten milliliters of an S. choleraesuis culture containing about 2.0 × 10⁸ cells per ml in the logarithmic growth phase was added to 50 ml of raw sewage and 0.2 ml of the same culture was added to 10 g of sludge for thermal treatment. The high concentration of Salmonella added to each sample, which exceeded the concentration of naturally occurring lactose-negative microorganisms, allowed direct determination of the Salmonella content by detection of yellow colonies on MacConkey agar.

Ten-milliliter portions of various phage suspensions containing approximately 10⁸ PFU of each phage studied ml⁻¹, which significantly exceeded the concentrations of the naturally occurring bacteriophages present in the sewage sample (Fig. 1), were added to 50-ml portions of sewage.

FIG. 1.

Box plots for the microbial load (log₁₀ units per 100 g or 100 ml) of the sludge (A) and the raw sewage (B) studied. Fecal coliforms (E. coli), fecal streptococci (FS), spores of SRC, somatic coliphages (SOMCPH), RNA F-specific phages (FRNAPH), and phages infecting B. fragilis strain RYC2056 (BFRPH) were examined. The values are the averages of nine experiments for the sludges and of six experiments for the raw sewage performed in duplicate.

One-milliliter portions of various virus suspensions containing approximately 10⁸ PFU of each enterovirus studied ml⁻¹ were added to 50-ml portions of sewage.

Thermal treatments. (i) Sludge.

Ten grams of sludge was placed into the bottom of a petri dish and incubated in an oven at 80°C. Samples were removed and tested to determine the numbers of microorganisms at the times indicated below.

(ii) Raw sewage.

Fifty-milliliter portions of raw sewage were introduced into 50-ml tubes and placed into a water bath at 60°C. Samples were removed and tested to determine numbers of microorganisms at the times indicated below.

Statistical analysis.

To obtain the descriptive statistics and to perform the Student's t and analysis of variance (ANOVA) tests, which were performed in order to determine the significance of the differences between results, the Statistical Package for Social Science (4) was used. Log₁₀-transformed values were used for all computations and tests. Log₁₀-transformed data were plotted as boxes and whiskers. This plot provided summary statistics for five numbers: the minimum, the maximum, the median, the 25th percentile, and the 75th percentile. Log₁₀ reduction values were calculated as follows: log₁₀ N_t − log₁₀ N_0, where N_t is the value at time t and N₀ is the value at the beginning of the experiment. Differences were considered significant at P < 0.05, as determined by the appropriate comparative test.

RESULTS

Inactivation of naturally occurring microorganisms in sludge.

The microbial load of the sludge used in the experiments described in this paper is summarized in Fig. 1. The mean values for the different indicators per 100 g of sludge ranged from 6.5 log₁₀ units for E. coli to 3.8 log₁₀ units for phages infecting B. fragilis.

The log₁₀ reductions in the numbers of naturally occurring phages and bacterial indicators in sludge after 30 min of incubation at 80°C were significantly different from the reductions after 90 min (P < 0.05, as determined by ANOVA) (Table 1). After 30 min of incubation at 80°C, the differences in inactivation between E. coli, the most heat-sensitive indicator, and phages infecting B. fragilis, somatic coliphages, and spores of sulfite-reducing bacteria were more than 3 log₁₀ units. The log₁₀ reductions exhibited by the latter three indicators did not differ significantly (P > 0.05, as determined by ANOVA). The results obtained for fecal streptococci and F-specific RNA bacteriophages were intermediate, and the log₁₀ reductions were greater than the log₁₀ reductions for the most resistant indicators by at least 1 log₁₀ unit

TABLE 1.

Log₁₀ reductions in the numbers of naturally occurring microorganisms after thermal treatment (80°C) of sludge^a

Organism	Log10 reduction after exposure for:
	30 min	90 min
E. coli	>3.6	>3.6
Fecal streptococci	>1.4, <1.8b	>2.7
SRC	0.3 (0.2)	0.4 (0.3)
Somatic coliphages	0.6 (0.1)	3.0 (0.7)
F-specific RNA phages	1.3 (0.1)	>3.4
Phages infecting B. fragilis	0.3 (0.1)	>2.7

^aUnless indicated otherwise, the values are averages for nine experiments performed in duplicate. The values in parentheses are standard deviations.

^bThe highest value when the values below the detection limit were considered zero and the lowest value when the values below the detection limit were considered the detection limit.

After 90 min of incubation at 80°C, only the log₁₀ reductions in the numbers of spores of SRC and somatic coliphages could be determined. In this case the values differed significantly (P < 0.05, as determined by Student's t test).

Inactivation of naturally occurring microorganisms in raw sewage.

The microbial load of the raw sewage used in the experiments reported in this paper is summarized in Fig. 1. The values for the various indicators studied were similar to the values described elsewhere (15).

The log₁₀ reductions in the numbers of naturally occurring phages, bacterial indicators, and enteroviruses after incubation at 60°C for 30 min were significantly different (Table 2). The inactivation of fecal coliform bacteria and E. coli was significantly greater (P < 0.05, as determined by ANOVA) than the inactivation of the other indicators. The fecal streptococci, whose level of inactivation differed significantly (P < 0.05, as determined by ANOVA) from the level of inactivation of phages, had intermediate resistance. Next in terms of sensitivity were the F-specific RNA phages, followed by the somatic coliphages, the phages infecting B. fragilis, and spores of SRC. The log₁₀ reductions in the numbers of B. fragilis and spores of SRC did not differ significantly (P > 0.05, as determined by Student's t test). The inactivation of the enteroviruses, although not quantifiable because of the low initial numbers, was clearly greater than the inactivation of the somatic coliphages, the phages infecting B. fragilis, and spores of SRC.

TABLE 2.

Log₁₀ reductions in the numbers of naturally occurring microorganisms after thermal treatment (60°C) of raw sewage^a

Organisms	Avg log10 reduction (SD)
Fecal coliforms (mFC method)	>5.6
Fecal coliforms (Fluorocult method)	6.2 (0.1)
E. coli	6.0 (0.3)
Fecal streptococci	3.4 (1.5)
SRC	0.1 (0.1)
Somatic coliphages	0.8 (0.2)
F-specific RNA phages	2.1 (0.3)
Phages infecting B. fragilis RYC2056	0.4 (0.1)
Culturable enteroviruses	>1.7

^aThe values are the averages for six experiments performed in duplicate.

Inactivation of phages and viruses added to sewage.

Various phages and viruses were added to raw sewage, and the levels of inactivation after 30 min of incubation at 60°C were determined (Table 3). There were significant differences between the levels of inactivation of some of the phages, although the differences never exceeded 2.5 log₁₀ units, which is low compared to the differences observed between naturally occurring phages and bacteria. Somatic coliphages SS13 and SC12 and phage B40-8 infecting B. fragilis were more resistant, and the log₁₀ reductions in the numbers of these phages differed significantly (P < 0.05, as determined by ANOVA) from the log₁₀ reductions in the numbers of the other phages, which also differed significantly (P < 0.05, as determined by ANOVA) from each other; phage MS2 was the least resistant. The six enteroviruses tested were significantly more sensitive (P < 0.05, as determined by ANOVA) than any of the phages. When we compared the levels of inactivation of the various enteroviruses, the log₁₀ reduction in the number of the most heat sensitive, the vaccine strain of poliovirus 1, differed significantly (P < 0.05, as determined by ANOVA) from the log₁₀ reductions in the numbers of the most heat-resistant viruses, which were coxsackievirus B5, echovirus 6, EV 1, and EV 2.

TABLE 3.

Log₁₀ reductions in the numbers of bacteriophages and viruses added to raw sewage after thermal treatment (60°C, 30 min)^a

Organisms	Avg log10 reduction (SD)
Bacteriophages
MY2	0.5 (0.1)
φX174	2.1 (0.2)
SR51	1.0 (0.1)
SC12	0.5 (0.1)
SS13	0.3 (0.1)
MS2	2.8 (0.1)
B40-8	0.4 (0.1)
Culturable enteroviruses
Poliovirus 1	5.4 (0.4)
Coxsackievirus B4	5.1 (0.6)
Coxsackievirus B5	4.8 (0.4)
Echovirus 6	4.3 (0.6)
EV 1	4.4 (0.5)
EV 2	4.3 (0.8)

^aThe values are the averages for six experiments performed in duplicate.

Inactivation of naturally occurring somatic coliphages and added Salmonella.

S. choleraesuis was added to the raw sewage and sludge, and the level of inactivation of this organism by thermal treatment for different times was compared to the levels of inactivation of naturally occurring somatic coliphages (Fig. 2). The level of inactivation of Salmonella was significantly greater (P < 0.05, as determined by Student's t test) than the level of inactivation of somatic coliphages; the reduction in the number of the somatic coliphages was quantifiable for much longer times than the time required for an 8-log₁₀ unit reduction in the number of added Salmonella cells.

FIG. 2.

Log₁₀ reductions in the levels of added S. cholerasuis (•) and naturally occurring somatic coliphages (□) in sludge after thermal treatment at 80°C (A) and in raw sewage after thermal treatment at 60°C (B) for different times.

DISCUSSION

Clear differences in resistance to thermal treatment of sludge between naturally occurring bacteria and phages were observed. The differences were significant and reflect a wide range of resistance, from E. coli, the most heat-sensitive indicator, to phages of B. fragilis and spores of SRC, the most heat-resistant indicators. All groups of phages were significantly more resistant than the bacterial indicators used, with the exception of spores of SRC. The lack of reduction or very low reduction in the numbers of spores of SRC indicates that these spores cannot be used to evaluate and validate thermal treatment processes for sludges. The ranking of the resistance of the naturally occurring microorganisms to thermal treatment was the same in raw sewage and dewatered sludge for different temperatures and times. Since raw sewage was the wettest possible sludge and the dewatered sludge used here was one of the driest sludges and since 80 and 60°C are the highest and lowest temperatures used for pasteurization of sludges, similar behaviors of the naturally occurring bacterial indicators and phages in a wide variety of sludges can be expected when the sludges are subjected to different thermal treatments.

Phages inoculated into sewage exhibited different levels of resistance to heat treatment, but the data were in agreement with the resistance data for naturally occurring bacteriophages. The levels of inactivation of phages MS2 and B40-8 were within the range of the levels of inactivation of naturally occurring F-specific RNA phages and phages infecting B. fragilis. Both groups of phages have been described as homogeneous groups (14, 19). Somatic coliphages exhibited significant differences. However, the Myoviridae and Siphoviridae are the most abundant groups in sewage (23), and the resistance to heat of phages like MY2, a member of the Myoviridae, and SR51, a member of the Siphoviridae, explains the low level of inactivation of naturally occurring somatic coliphages. These differences indicate that the results of this sort of experiment performed with a single phage cannot be extrapolated to all phages.

Although somatic coliphages have a wide range of sensitivities to heat, they may mimic what happens with human viruses. Indeed, very significant differences in the resistance to heat of viruses potentially present in sludge have been reported (8, 22, 29). The comparison of the levels of resistance of enteroviruses and phages reported here, a comparison of the levels of resistance of enteroviruses and parvoviruses with the level of resistance of phage f2, an F-specific RNA phage (29), and a comparison of the levels of resistance of phages MS2 and φX174 with the level of resistance of hepatitis A virus (22) suggest that most human viruses are more readily inactivated by heat treatment than naturally occurring somatic coliphages and phages infecting B. fragilis are. Considering the previously described levels of resistance and the previously reported numbers of human viruses in sludges (7, 21, 28, 31, 38), heat treatments that reduce the numbers of somatic coliphages and phages infecting B. fragilis between 0.5 and 1.0 log₁₀ unit should guarantee reductions in the numbers of infectious human viruses to levels below the levels required by some guidelines (34, 35).

Comparative data on the effects of thermal treatment on bacteriophages and ova and oocysts of parasites are not available. However, the previously described effects of heat on the infectivity of ova and oocysts (26, 37) of parasites indicate that the infectivity is clearly reduced by thermal treatment. Thus, heat treatments that reduce the numbers of somatic coliphages and bacteriophages infecting B. fragilis between 0.5 and 1.0 log₁₀ unit should guarantee reductions in the numbers of infectious oocysts of parasites to levels below the levels required by some guidelines (34, 35).

The results presented here suggest that after calibration studies it should be possible to determine the ranges for inactivation of naturally occurring bacteriophages that guarantee a range of reduction in the number of Salmonella cells. Additionally, for the three groups of phages studied the numbers in sludges were high enough so that the phages could be counted after thermal treatments that reduced the numbers of E. coli and Salmonella cells by more than 6 log₁₀ units, and the numbers of somatic coliphages were high enough so that the phages in the treated sludge could be counted after the number of Salmonella cells was reduced by more than 8 log₁₀ units.

In our opinion data presented here indicate that naturally occurring bacteriophages are a useful tool for evaluating the effect of thermal treatment of sludges on a wide range of microorganisms. Indeed, the numbers of these bacteriophages in sludges are high enough so that they can be counted in 1 g or at most 10 g of sludge after a simple extraction procedure. For the group of phages that seems to be most useful, we recommend somatic coliphages because of the high numbers of these phages in sludges and because the method used for detection is very easy, cheap, and fast.