ARTICLE IN PRESS

Int. J. Hyg. Environ.-Health 209 (2006) 103–107 www.elsevier.de/ijheh

Isolation of Salmonella serotypes in wastewater and effluent: Effect of treatment and potential risk Elena Espigares, Aurora Bueno, Miguel Espigares, Ramo´n Ga´lvez Department of Preventive Medicine and Public Health, University of Granada, 18071 Granada, Spain Received 22 February 2005; received in revised form 7 August 2005; accepted 30 August 2005

Abstract The characteristics of strains of Salmonella isolated in wastewater were compared before and after treatment with activated sludge (conventional sewage treatment). Of the 21 strains studied, 11 came from raw wastewater and 10 from treated water. Our objective was to determine if the treatment affected all serotypes to the same extent, and whether it had an effect on the plasmidic composition and selection of resistances to antibiotics and disinfectants. The serotypes of all strains were identified, and we also determined the sensitivity to antibiotics, the minimal bactericidal concentration of three disinfectants (peracetic acid, glutaraldehyde, and sodium hypochlorite), and plasmidic content. Results showed no significant differences in the isolated strains before versus after treatment, thus indicating that the conventional treatment of sewage does not constitute a risk factor in the selection of Salmonella strains with a greater pathogenic potential. r 2005 Elsevier GmbH. All rights reserved. Keywords: Antibiotic resistance; Disinfectant; Plasmid; Salmonella; Wastewater

Introduction Foods and water still play the main roles in the transmission of Salmonella, though wastewater is particularly important. Surface water is often contaminated by urban wastewater, by effluents of meat industries and by wastewater from livestock ranches, involving different serotypes of Salmonella (Rusin et al., 2000). The serotypes isolated from human samples do not always coincide with the serotypes isolated from wastewater, yet this is not always the case in Spain, where Salmonella enteritidis (50.7%), S. typhimurium (23.2%), and S. hadar (4.7%) were the most frequent serotypes isolated from clinical Corresponding author.

E-mail address: [email protected] (M. Espigares). 1438-4639/$ - see front matter r 2005 Elsevier GmbH. All rights reserved. doi:10.1016/j.ijheh.2005.08.006

human samples in the year 2000, and together with S. anatum were also the ones most often isolated from water samples (Usera et al., 2001). Although wastewater and drinking water are treated to eliminate pathogenic microorganisms and prevent waterborne transmission, numerous studies indicate that conventional wastewater treatment does not guarantee their complete elimination. Our research group recently found treated water to contain a Salmonella MPN of 45/ 100 ml (Howard et al., 2004). The survival of Salmonella despite treatment implies the possibility of selection of the most resistant strains, or the acquisition of resistance through the transference of genetic material. Thus a double public health risk may exist: the presence of Salmonella itself in the effluent, and the predominance of more pathogenic strains.

ARTICLE IN PRESS 104

E. Espigares et al. / Int. J. Hyg. Environ.-Health 209 (2006) 103–107

We studied strains of Salmonella isolated in wastewater to observe the effect of sanitary treatment on the selection of serotypes and resistances to a number of antiobiotics and disinfectants.

Material and methods Bacterial strains This study involved 21 strains of Salmonella isolated from 42 samples obtained at the wastewater treatment plant of Granada (Southern Spain) during the period November 1999–March 2000. The samples were obtained within a 24 h period, with two samples taken simultaneously: one of raw wastewater and the other of the effluent after aerobic biological treatment. From the 21 samples of raw wastewater, 11 strains were isolated (52.4% of positive samples), and 10 strains proceeded from the 21 samples (47.6% positive) of the effluent that had undergone activated sludge treatment (Howard et al., 2004). Salmonella strains were isolated using the selective Rappaport–Vassiliadis medium (Oxoid) and Salmonella–Shigella agar (Oxoid). Biochemical identification was performed with analysis of serotypes based on slide seroagglutination using somatic and flagelar Salmonella antisera (Centro Nacional de Microbiologı´ a del Instituto de Salud Carlos III, Majadahonda, Madrid, Spain). The following reference strains were used: Pseudomonas aeruginosa CIP A22 and Escherichia coli CIP 54127 for disinfectant activity testing; and E. coli V517 as the marker of molecular weight of plasmids.

Antibiotic susceptibility testing Antibiotic susceptibility was determined by the agar diffusion method, initially described by Kirby and Bauer, and standardized by the US Federal Drug Administration (FDA) and by the World Health Organization (Acar and Goldstein, 1996). The following antimicrobial susceptibility test disks were used: amikacine (An), amoxicillin/clavulanic acid (AmC), ampicillin (Ap), azthreonam (Atm), cefepime (Fep), cefotaxime (Ctx), cefoxitin (Fox), cefuroxime (Cxm), cephalothin (Cf), ciprofloxacin (Cip), chloramphenicol (C), fosfomycin (Fos), gentamicin (Gm), imipenem (Ipm), nalidixic acid (Na), neomycin (N), nitrofurantoin (Fm), piperacillin (Pip), sulfisoxazole (Su), tetracycline (Te), and trimethoprim/sulfamethoxazole (Sxt).

Disinfectant activity testing Three high level commercial disinfectants were used: 2% glutaraldehyde (Instrunets), 0.26% peracetic acid

(Perasafes), and sodium hypochlorite (liquid bleach). The available chlorine in sodium hypochlorite solutions was determined using the iodometric method (APHA, 1989). Disinfectant activity was tested in accordance with the norms UNE-EN 1040 and AFNOR NF T 72-150 for the evaluation of bactericidal activity by means of the dilution-neutralization method (AFNOR, 1995; European Standardization Committee, 1997; A`lvarez Alca´ntara et al., 2001). The temperature of the assay was 20 1C and the time of contact was 5 min. From the concentration of use, a series of 1/2 dilutions of the disinfectants was tested to determine the minimal bactericidal concentration (MBC). The composition of 1000 ml of the neutralizer at a double concentration (
2) was as follows: Tween 80, 129 g (120 ml); 40% sodium bisulphite, 25 ml; sodium thiosulphate pentahydrate, 15.69 g; adjusted to pH 7.0 and sterilized by filtration (Espigares et al., 2003).

Plasmid analysis Plasmids were isolated from the bacterial strains according the Quantum Prep kit instructions (Bio-Rad). The isolated plasmids were characterized by 0.7% agarose gel electrophoresis (Maniatis et al., 1989). The molecular weights of the plasmids were obtained by comparing their relative mobilities to standard DNA markers.

Statistical analysis The statistical treatment of the data was done with the SPSS for Windows software package (SPSS Inc., Chicago, USA). After basic statistical analysis of data, the Student’s t-test was applied for two independent samples.

Results and discussion Because this study involves urban wastewater, it was to be expected that the strains of Salmonella isolated would coincide with those most frequently isolated from humans over the same geographical area. Data published for Spain show that the four most frequent serotypes in clinical samples of human origin within the year 1999 were: 46.1% of S. enteritidis, 20.2% of S. typhimurium, 8% of S. hadar, 4.6% of S. 4,5,12:i:-(I), and 3.0% of S. virchow (Usera et al., 2000). Similarly, for the period 1999–2000 the serotypes most frequently isolated in Europe were S. enteritidis, S. typhimurium, S. hadar, and S. virchow (Fisher, 2000), while in the United States S. enteritidis and S. typhimurium comprised 42%

ARTICLE IN PRESS E. Espigares et al. / Int. J. Hyg. Environ.-Health 209 (2006) 103–107

105

Table 1. Salmonella strains isolated from raw and treated wastewater

Table 2. Antibiotic resistance among Salmonella strains isolated from raw and treated wastewater

Serotypes

Number of isolations

Strain

Resistances

From raw water From treated water Total

Raw wastewater S. enteritidis HFG 56 S. goldcoast HFG 23 S. london HFG 21 S. london HFG 59 S. newport HFG 53 S. anatum HFG 44 S. enteritidis HFG 34 S. hadar HFG 47 S. typhimurium HFG 50 S. hadar HFG 26 S. hadar HFG 39

Susceptible Susceptible Susceptible Susceptible Susceptible Ap, Fm, Te, Sxt AmC, Ap, Pip Cf, Pip, Na, Te AmC, Ap, C, Pip, Te AmC, Ap, Cf, Na, Pip, Te AmC, Ap, Cf, Na, Pip, Te

Treated wastewater S. enteritidis HFG 40 S. enteritidis HFG 46 S. london HFG 43 S. enteritidis HFG 52 S. hadar HFG 22 S. hadar HFG 38 S. anatum HFG 35 S. hadar HFG 58 S. hadar HFG 33 S. hadar HFG 55

Susceptible Susceptible Susceptible Fm Na, Te Na, Su Ap, Su, Te, Sxt AmC, Ap, Cf, Na, Pip AmC, Ap, Cf, Na, Pip, Te AmC, Ap, Cf, N, Pip, Te

S. S. S. S. S. S. S.

anatum enteritidis goldcoast hadar london newport typhimurium

Total

1 2 1 3 2 1 1

1 3 — 5 1 — —

2 5 1 8 3 1 1

11

10

21

No statistically significant differences between serotypes isolated from raw and treated waters (independent samples t-test, po0:05).

of all isolates (CDC, 2000). These findings suggest characteristic patterns for developed nations worldwide. The most frequent serotypes isolated in our samples come to support the above findings, though some exceptions are noteworthy (Table 1). For instance, the most frequent serotype was S. hadar (38.1%), followed by S. enteritidis (23.8%), S. london (14.3%), and S. anatum (9.5%). The least frequent serotypes were S. typhimurium, S. goldcoast, and S. newport, with one isolate each (4.8%). In 1999 and 2000 in Spain the most frequent serotypes isolated from wastewater were S. enteritidis and S. anatum (Usera et al., 2001). The existence of differences between serotypes proceeding from wastewater and from humans is a common finding as wastewater can often contain strains of animal origin. Epidemiologically speaking, one very important matter is the presence of Salmonella strains in treated water. Previously published data (Howard et al., 2004), show that municipal wastewater having undergone an activated sludge process continued to bear Salmonella: the raw water yielded an MPN of 266.7/100 ml and the treated water 45/100 ml, representing a reduction of only 83%. This reveals a considerable risk of the transmission of salmonellosis. Out of the 21 isolates of our study, eight strains (31.8%) were found to be susceptible to all the antibiotics tested. The strain S. enteritidis HFG 52 presented resistance exclusively to nitrofurantoin, S. hadar HFG 38 showed a double resistance to nalidixic acid and sulfisoxazole, and S. hadar HFG 22 was resistant to nalidixic acid and tetracycline (Table 2). The rest presented multiple antibiotic resistance. Other authors describe the increasing isolation of strains with antibiotic multiresistance (Hald et al., 2003; Van Pelt et al., 2003). We should underline that 42.9% were resistant to at least four antimicrobial compounds and four strains of S. hadar exhibited a pattern of multirresistance to six antibiotics: three were resistant to AmC, Ap, Cf, Na, Pip, and Te, and the fourth was

Amikacine (An), amoxicillin/clavulanic acid (AmC), ampicillin (Ap), azthreonam (Atm), cephalothin (Cf), cefepime (Fep), cefotaxime (Ctx), cefoxitin (Fox), cefuroxime (Cxm), ciprofloxacin (Cip), chloramphenicol (C), fosfomycin (Fos), gentamicin (Gm), imipenem (Ipm), nalidixic acid (Na), neomycin (N), nitrofurantoin (Fm), piperacillin (Pip), sulfisoxazole (Su), tetracycline (Te), and trimethoprim/sulfamethoxazole (Sxt).

resistant to neomycin instead of nalidixic acid, though likewise resistant to AmC, Ap, Cf, Pip, and Te. When antibiotics are considered individually, tetracycline and ampicillin – both of great importance in the treatment of gastrointestinal infections – show the highest percentage of resistant strains (42.9%). This percentage is clearly on the rise, as a study we undertook in 1984 involving 385 strains showed 15.8% to be resistant to tetracyline and only 8.3% to be resistant to ampicillin, with many of the resistances being of extrachromosomic origin (Parras Jime´nez et al., 1984); this study found all the strains to be susceptible to cephalothin, whereas the results presented here indicate a resistance of 28.8%. Salmonella is easily transmitted by water, and so water disinfection constitutes a key preventive mechanism. All strains tested were susceptible to the concentration of use of glutaraldehyde (2%; 20,000 mg/ml) and 0.26% peracetic acid (Table 3). This is very important because trials in pilot-plants using peracetic acid to disinfect water have given very positive results (Veschetti et al., 2003).

ARTICLE IN PRESS 106

Table 3. isolated

E. Espigares et al. / Int. J. Hyg. Environ.-Health 209 (2006) 103–107

Minimal bactericidal concentration (MBC) of three high level commercial disinfectants used for the Salmonella strains

Type of water

Isolates

Glutaraldehyde (mg/ml) Mean7SD

Peracetic acid (mg/ml) Mean7SD

Sodium hypochlorite (mg/ml) Mean7SD

Untreated Treated

11 10

664.77227.5 618.77177.9

11.175.0 12.776.9

33.879.2 41.0714.4

No statistically significant differences between untreated and treated waters (independent samples t-test, po0:05).

Table 4. Number of plasmids, molecular weight and plasmidic total among Salmonella strains isolated from raw and treated wastewater Type of water

Isolates

Average number of plasmids (range)

Average molecular weights (range)

Average total kb (range)

Untreated

11

Treated

10

2.81 (0–9) 3.90 (0–9)

8.09 (2.0–49.3) 9.74 (2.0–49.0)

22.8 (0–59.1) 36.3 (0–95.5)

Plasmidic total is the sum of the molecular weights of all the plasmids in each strain.

Sodium hypochlorite is widely used in water chlorination. When considering a concentration of use of 1%, the MCB of all the strains is lower than the dilution of 1/ 128 (Table 3). An important question is whether wastewater treatment contributes to the selection process of resistant strains. Treatment is carried out in two stages: primary sedimentation and the aerobic activated sludge process. Both these processes reduce the concentrations of pathogenic microorganisms. In the biological treatment a complex ecosystem is created, with the participation of bacteriophages, bacteria, protozoa, algae, and other organisms. In this ecosystem, water undergoes a decrease in organic contents, and the concentration of pathogens is reduced by antagonistic microorganisms as well as adsorption to or incorporation into the secondary sludge. Conventional wastewater treatment decreases the number of pathogenic microorganisms, but does not completely eliminate them. The coexistence and competition among strains, species and even different genera can also allow interchanges of genetic material, and favor the selection of strains that resist the aformentioned processes of antibiosis. Wastewater treatment might have a potential impact on: (a) the survival of the different serotypes; (b) increased resistance to antibiotics and disinfectants; and (c) changes in the contents of extrachromosomic genetic material. In opposition to our initial hypothesis, we did not observe a statistically significant alteration in the composition of serotypes from raw wastewater and treated effluent (Table 1), meaning the two different serogroups have a similar degree of resistance to the environmental conditions.

No strains were found to be resistant to the disinfectants tested, and the minor differences in the susceptibility of raw versus treated waters were not statistically significant (Table 3). Nor did we obtain statistically significant differences for the respective antibiotic resistance of the strains proceeding from raw wastewater and treated effluent. The data given in Table 2 point to a mean value of 2.55 resistances to antibiotics per strain proceeding from raw wastewater; and 2.20 resistances per strain from treated water. The slightly lower resistance of the strains from treated water is not statistically significant (p ¼ 0:761), and is similar to the findings of previous authors studying fecal coliforms (Hassani et al., 1999). Because many processes of resistance are codified by extrachromosomic genetic material, the analysis of plasmidic contents is of great interest (see Table 4). The strains of Salmonella isolated presented a mean value of 3.33 plasmids/strain (standard deviation, 3.32), with values of 2.8273.43 and 3.9073.28 for the strains proceeding from raw wastewater and treated water, respectively. Again, these differences are not statistically significant (p ¼ 0:470). Another parameter, total plasmidic content, was obtained by adding up the molecular weights of all the plasmids in a strain. We arrived at a mean value of 29.23728.07 kb/strain, with respective values of 22.81726.23 and 36.30729.68 kb/strain for the raw and treated water, a difference that was not statistically significant (p ¼ 0:283). The findings described here would indicate that aerobic treatment using activated sludge has no real influence on the selection of serotypes and resistances to antibiotics and disinfectants; and while it appears to

ARTICLE IN PRESS E. Espigares et al. / Int. J. Hyg. Environ.-Health 209 (2006) 103–107

influence the increase in plasmidic contents, this result is not significant. This finding is of great epidemiological importance, as it allows us to conclude that the biological treatment of wastewater does not constitute a risk factor in the selection of pathogenic strains of Salmonella.

Acknowledgments The authors would like to thank Dr. Echeita and Dr. Usera of the Bacteriological Section of the Centro Nacional de Microbiologı´ a del Instituto de Salud Carlos III (Madrid, Spain) for their help in the serotypification of the Salmonella strains, and Jose´ Luis Martı´ n Ga´lvez, the Director of EDAR-EMASAGRA, for his cooperation and technical support in the collection of samples. Moreover we thank Ce´sar Criado Sa´nchez for his capable assistance in our laboratory, and Jean Sanders for her helpful suggestions while translating and editing the manuscript.

References Acar, J.F., Goldstein, F.W., 1996. Disk susceptibility test. In: Lorian, V. (Ed.), Antibiotics in Laboratory Medicine. Williams & Wilkins, Baltimore, MD, pp. 1–51. AFNOR, 1995. Antiseptiques et disinfectants, third ed. AFNOR, Paris. A`lvarez Alca´ntara, A., Espigares Rodrı´ guez, E., Ga´lvez Vargas, R., 2001. Valoracio´n de desinfectantes. Me´todo de dilucio´n-neutralizacio´n. Hig. Sanid. Ambient. 1, 1–5. APHA, 1989. Standard Methods for Examination of Water and Wastewater, 17th ed. American Public Health Association, Washington, DC, pp. 4, 48–51. CDC, 2000. Summary of notifiable diseases – United States, 2000. MMWR 49, 1–102. Espigares, E., Bueno, A., Ferna´ndez-Crehuet, M., Espigares, M., 2003. Efficacy of some neutralizers in suspensio´n tests determining the activity of disinfectants. J. Hosp. Infect. 55, 137–140. European Standardization Committee, 1997. Norma espan˜ola UNE-EN 1040. Antise´pticos y desinfectantes quı´ micos. Actividad bactericida ba´sica. Me´todo de ensayo y requisitos (fase 1). AENOR, Madrid. Fisher, I., Salmonella in Europe – enter-net report. Eurosurveillance 4, 7/09/2000. Hald, T., Wingstrand, A., Swanenburg, M., von Altrock, A., Thorberg, B.M., 2003. The occurrence and epidemiology of

107

Salmonella in European pig slaughterhouses. Epidemiol. Infect. 131, 1187–1203. Hassani, L., Rafouk, L., Aı¨ t Alla, A., 1999. Antibiotic resistance among faecal coliform bacteria isolated from wastewater before and after treatment by an experimental sand filter. World J. Microbiol. Biotechnol. 15, 317–319. Howard, I., Espigares, E., Lardelli, P., Martı´ n, J.L., Espigares, M., 2004. Evaluation of microbiological and physicochemical indicators for wastewater treatment. Environ. Toxicol. 19, 241–249. Maniatis, T., Fritsch, E.F., Sambrook, J., 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, New York, pp. 150–163. Parras Jime´nez, L., Espigares Garcı´ a, M., Rodrı´ guez-Contreras, R., A`lvarez Alca´ntara, A., Ga´lvez Vargas, R., 1984. Resistencia de origen plasmı´ dico a los b-lacta´micos en el ge´nero Salmonella. Laboratorio 78, 31–41. Rusin, P., Enriquez, C.E., Jonson, D., Yerba, C.P., 2000. Environmentally transmitted pathogens. In: Maier, R.M., Pepper, I.L., Gerba, C.P. (Eds.), Environmental Microbiology. Academic Press, San Diego, pp. 447–489. Usera, M.A., Aladuen˜a, A., Diaz, R., De La Fuente, M., Cerda´n, P., Gutierrez, R., Echeita, A., 2000. Ana´lisis de las cepas de Salmonella spp aisladas de muestras clı´ nicas de origen no humano en Espan˜a en el an˜o 1999. Bol. Epidemiol. Semanal 8, 133–144. Usera, M.A., Aladuen˜a, A., Dı´ ez, R., De La Fuente, M., Gutie´rrez, R., Cerda´n, P., Echeita, A., 2000. Ana´lisis de las cepas de Salmonella spp aisladas de muestras clı´ nicas de origen humano en Espan˜a en el an˜o 1999 (I). Bol. Epidemiol. Semanal 8, 45–52. Usera, M.A., Aladuen˜a, A., Dı´ az, R., De La Fuente, M., Cerda´n, P., Gutie´rrez, R., Echeita, A., 2001. Ana´lisis de las cepas de Salmonella spp aisladas de muestras de origen no humano en Espan˜a en el an˜o 2000. Bol. Epidemiol. Semanal 9, 281–292. Usera, M.A., Aladuen˜a, A., Diez, R., De La Fuente, M., Gutie´rrez, R., Cerda´n, P., Arroyo, M., Gonza´lez, R., Echeita, A., 2001. Ana´lisis de las cepas de Salmonella spp aisladas de muestras clı´ nicas de origen humano en Espan˜a del 2000 (I). Bol. Epidemiol. Semanal 9, 221–228. Van Pelt, W., de Wit, M.A.S., Wannet, W.J.B., Ligtvoet, E.J.J., Widdowson, M.A., van Duynhoven, Y.T.H.P., 2003. Laboratory surveillance of bacterial gastroenteric pathogens in The Netherlands, 1991–2001. Epidemiol. Infect. 130, 431–441. Veschetti, E., Cutilli, D., Bonadonna, L., Briancesco, R., Martini, C., Cecchini, G., Anastasi, P., Ottaviani, M., 2003. Pilot-plant comparative study of peracetic acid and sodium hypochlorite wastewater disinfection. Water Res. 37, 78–94.