Abstract
Molecular epidemiology at local scale in Sicily (Italy) of ampicillin resistance in Salmonella spp. isolates from municipal wastewater (n = 64) and clinical specimens (n = 274) is described in comparison with previously examined Escherichia coli isolates (n = 273) from wastewater. High prevalence of antibiotic resistance (28.9%) with highest resistance rates against ampicillin (22.7%) was observed in E. coli isolates. Different resistance rates were observed in Salmonella according to the serovars, with prevalences of the same order in both wastewater and clinical isolates belonging to the same serovar (e.g., 91.7% ampicillin resistance in wastewater isolates vs. 70.8% in clinical isolates of the Salmonella serovar Typhimurium and 0% ampicillin resistance in both wastewater and clinical isolates of the Salmonella serovar Enteritidis). The β-lactam resistance gene bla TEM was present in both wastewater and clinical Salmonella spp. isolates, with the exception of Salmonella enterica serovar Typhimurium isolates with a typical six-drug resistance pattern AmpChlSulTeStrSp that had the bla PSE-1 gene. The bla TEM gene was present in all the E. coli isolates but one had the bla SHV gene. Several E. coli and some Salmonella isolates were positive for class 1 integrons with variable regions of 1.0 or 1.5 kb containing aadA1, dfrA17-aadA5, or dfrA1-aadA1 gene cassettes, whereas Salmonella serovar Typhimurium isolates with the six-drug resistance pattern were positive for both 1.0 and 1.2 kb integrons. Polymerase chain reaction replicon typing demonstrated the presence of multireplicon resistance plasmids in several isolates of E. coli, containing two to four of the replicons IncF, IncI1, IncFIA, and IncFIB, whereas other isolates showed resistance plasmids with only IncF, IncP, or IncK replicons. Replicon IncI1 was detected in one Salmonella isolate, whereas other isolates belonging to different serovars had IncN replicons. Analysis of isolates from wastewater can be a useful epidemiologic tool to monitor the prevalence of antibiotic resistance and genetic elements related to antibiotic resistance in Salmonella clones circulating in the human population.
Introduction
D
Commensal enteric bacteria are known to constitute reservoirs of genes determining antibiotic resistance (Shehabi et al., 2006; Su et al., 2006). In particular, commensal Escherichia coli can transfer their resistances in the gastrointestinal environment, especially when antimicrobial exposure occurs, determining high prevalence of antibiotic resistance; when this occurs, mobile elements contribute significantly to the resistance spread (Smith et al., 2007). Resistance plasmids (R plasmids) and integrons have been implicated in this spread among the Gram-negative bacteria, especially enterobacteria, including various Salmonella serovars. The ability of bacteria to acquire and disseminate exogenous genes via mobile genetic elements such as plasmids and transposons has been the major factor in the development of multiple drug resistance over the last 50 years (Hawkey and Jones, 2009).
Molecular methods applied to the study of R plasmids and integrons are of capital importance to understand the epidemiology of antibiotic resistance among bacterial and human populations. However, understanding the epidemiological role of different genetic elements implicated in antibiotic resistance is complex because of the diversity and promiscuity of these elements (Lévesque et al., 1995). In some instance, molecular typing has been proved to be able to describe the dissemination and to follow the evolution of particular resistance plasmids at large scale (Carattoli et al., 2002; Hopkins et al., 2006). At local scale, the study of enterobacteria isolates from urban wastewater could be a useful tool to easily determine the range of antibiotic resistance in the bacterial populations colonizing the intestine of residents.
In a previous paper (Pignato et al., 2009) we have evaluated the prevalence of antibiotic resistance among E. coli isolates from raw and treated municipal wastewater used in agriculture, showing high rates of multiple resistance, including resistance against ampicillin. Several ampicillin-resistant isolates were able to transfer en bloc their resistance patterns by conjugative R plasmids and some were proved to carry class 1 integrons with gene cassettes of size 1.0 or 1.5 kb. On the basis of these results we concluded that antibiotic-resistant E. coli surviving treatment processes can act as a reservoir of antibiotic resistance genes, particularly ampicillin resistance, which can be spread to the pathogenic and commensal enteric bacteria in human and animal consumers of crop irrigated by wastewater.
As resistance against ampicillin was the most frequent resistance in the E. coli isolates previously examined, the objectives of this study were (i) to determine the prevalence of ampicillin resistance among Salmonella spp. isolates from municipal wastewater and clinical specimens obtained in the same geographic location during 1 year of monitoring; (ii) to detect and characterize bla genes, class 1 integrons, and replicon types in ampicillin-resistant Salmonella spp. and in E. coli isolated from the same wastewater effluents; and (iii) to investigate eventual relatedness among plasmids and integrons implicated in ampicillin resistance of the Salmonella spp. and E. coli isolates.
These objectives were aimed at investigating whether the study of isolates from wastewater can be a useful and easy to perform epidemiologic tool to monitor the prevalence of antibiotic resistance and genetic elements related to drug resistance in Salmonella clones circulating in the human population.
Materials and Methods
Isolation and identification of bacteria
Wastewater samples (108 samples in all, 36 from crude, 36 from treated, and 36 from refined wastewater) from two towns (Caltagirone and San Michele di Ganzaria) located in the district of Catania, in eastern Sicily (Italy), were collected from the respective municipal treatment plants and processed every 2 weeks (Plant A) or every month (Plant B) over a 1-year period, from January to December 2004, as previously described (Pignato et al., 2009). No wastes from farms or slaughterhouses inflowed to the urban wastewater that were treated in the two plants. E. coli isolates were those analyzed in a previous study (Pignato et al., 2009). They were isolated by standard membrane filtration technique (APHA AWWA WEF, 1995) using 47-mm acetate membrane filters (Sartorius AG, Goettingen, Germany) with a nominal pore size of 0.45 μm, which were placed on the surface of plates of CHROMagar E. coli (CHROMagar, Paris, France), a chromogenous selective medium. After 24 h of incubation at 37°C, plates were inspected for growth and two or three colonies per plate showing the typical morphological characteristics of E. coli were selected for biochemical identification by the API 20E system (bioMérieux, Marcy L'Etoile, France). Salmonella spp. was detected and quantified from wastewater by a most probable number procedure using Salmosyst broth base and Salmosyst selective supplement tablets (Merck, Darmstadt, Germany) as a liquid medium for preenrichment (6 h at 37°C in Salmosyst broth base) and selective enrichment (18 h at 37°C in the same tube of Salmosyst broth base after adding a tablet of selective supplement) and Rambach agar (Merck) for plating according to a procedure for rapid detection and isolation of these bacteria (Pignato et al., 1995). Red colonies that developed after 24 h of incubation at 37°C (five colonies from each plate) were submitted to biochemical identification by API 20E system and serotyping by polyvalent and monovalent anti-O and anti-H sera (either Bio-Rad, Marnes la Coquette, France, or Biogenetics, Ponte San Nicolò, Padova, Italy). One isolate of each serovar obtained from each positive sample was retained for the resistance study. During the same 1-year period, Salmonella spp. was also isolated from clinical specimens (feces) of patients hospitalized for gastroenteritis at the district hospital (Gravina Hospital) of one of the two towns (Caltagirone, CT) conferring its wastewater to the monitored plant A. Feces were processed by standard microbiological methods and plated on Rambach agar. Red colonies that developed on Rambach agar from wastewater or feces after 24 h of incubation at 37°C were submitted to biochemical identification as described earlier.
Isolates from wastewater
A total of 273 previously examined E. coli isolates and 64 Salmonella isolates from wastewater were studied. The following serovars were examined: S. enterica subsp. enterica serovars Typhimurium (n = 36), Enteritidis (n = 6), Blockley (n = 4), Napoli (n = 4), Newport (n = 3), Bovismorbificans (n = 2), Braenderup (n = 2), Corvallis (n = 1), Infantis (n = 1), Kenya (n = 1), Montevideo (n = 1), and Muenster (n = 1), S. enterica subsp. salamae serovar 6,7:b:z39 (n = 1), and S. enterica subsp. diarizonae serovar 61:r:z (n = 1).
Isolates from clinical specimens
A total of 274 Salmonella isolates from feces of patients suffering from gastroenteritis were studied. The following serovars were identified: S. enterica subsp. enterica serovars Typhimurium (n = 144), Enteritidis (n = 62), Blockley (n = 18), Napoli (n = 13), Muenchen (n = 9), Newport (n = 7), Goldcoast (n = 3), Heidelberg (n = 3), Bovismorbificans (n = 2), Braenderup (n = 2), Infantis (n = 2), Bredeney (n = 1), Derby (n = 1), Hadar (n = 1), Isangi (n = 1), Kenya (n = 1), Muenster (n = 1), and Panama (n = 1) and S. enterica subsp. salamae serovars 41:z10:1,2 (n = 1) and 42:b:e,n,x,z15 (n = 1).
Antibiotic susceptibility testing
Isolates were screened for resistance to eight antibiotics (Sigma-Aldrich, Steinheim, Germany) by streaking a loopfull of 18-h broth cultures on Mueller–Hinton agar (Oxoid, Hampshire, England) plates containing ampicillin (Amp) 32 μg/mL, chloramphenicol (Chl) 32 μg/mL, sulphamethoxazole (Sul) 512 μg/mL, tetracycline (Te) 16 μg/mL, trimethoprim 10 μg/mL, streptomycin (Str) 64 μg/mL, kanamycin 64 μg/mL, and nalidixic acid (Nal) 64 μg/mL. Plates were incubated for 24 h at 37°C and isolates that yielded bacterial growth were recorded as resistant to the corresponding antibiotic. The antibiotic concentrations were some above the recommended cutoff points. These concentrations were adopted to avoid ambiguous results. Therefore, the prevalence of resistant isolates might have been underestimated. A selected number of ampicillin-resistant isolates (36 Salmonella and 30 E. coli isolates) exhibiting representative antibiotic resistance patterns were tested by the disk diffusion method with 32 antimicrobial drugs (Bio-Rad). Isolates were categorized as susceptible, intermediate, or resistant, according to the Antibiogram Committee of the French Society for Microbiology cutoff values (
Resistance transfer determination
In total, 66 ampicillin-resistant isolates (30 E. coli, 8 Salmonella from wastewater, and 28 Salmonella from feces), representative of different antibiotic resistance patterns, were investigated for the transferability of their resistance traits to E. coli C1a (nalA) and J5 (rif). The donor and recipient strains were grown in TSB (Difco, Detroit, MI) to logarithmic phase and then mixed in equal volumes (1 mL + 1 mL) and incubated at 37°C for 18 h. Transconjugant clones were selected on Mueller–Hinton agar containing ampicillin (100 μg/mL) and nalidixic acid (64 μg/mL) or rifampicin (250 μg/mL). Transfer of the resistances was confirmed by testing the transconjugants against the antibiotics corresponding to the donor resistances.
PCR amplification and DNA sequencing
Total DNA was extracted using the InstaGene matrix kit (Bio-Rad) according to the manufacturer's recommendations. The ampicillin resistance genes, bla TEM, bla SHV, bla PSE-1, and bla OXA-1 group, and class 1 integron gene cassettes were amplified by polymerase chain reaction (PCR) as described previously (Lévesque et al., 1995; Weill et al., 2006).
Sequencing was performed at the “Plateforme de Génotypage des Pathogènes et Santé Publique, PF8” (Institut Pasteur, Paris, France). The nucleotide sequences and the deduced protein sequences were analyzed with EditSeq and Megalign software (Dnastar, Madison, WI). The BLASTN program of National Center for Biotechnology Information was used for database searches (
Replicon typing
PCR-based replicon typing was performed according to Carattoli et al. (2005) to type the R plasmids carried by 17 E. coli C1a transconjugants obtained from isolates of Salmonella and E. coli. Eighteen primer pairs were used to perform simplex PCRs, which recognized HI1, HI2, I1-IY, X, L/M, N, FIA, FIB, W, Y, P, FIC, A/C, T, FIIA, F, K, and B/O replicons.
Results
The numbers and percentages of Salmonella and E. coli isolates resistant to one or more antibiotics and also numbers and percentages of isolates resistant to ampicillin are reported in Table 1. Isolates of Salmonella showed different degrees of antibiotic resistance according to the serovars, with prevalences, including ampicillin resistance prevalence, of the same order in isolates from wastewater and in isolates from clinical specimens. Also, resistance profiles were similar in isolates from wastewater and clinical specimens according to the serovar. In particular, the six-drug resistance pattern AmpChlSulTeStrSp was observed with a similar order of frequency in Salmonella serovar Typhimurium isolates from wastewater and clinical specimens, as further specified below.
Isolates resistant to one or more antibiotics.
S. enterica subsp. enterica serovars Bovismorbificans, Braenderup, Corvallis, Infantis, Kenya, Montevideo, and Muenster and S. enterica subsp. diarizonae serovar 61:r:z.
S. enterica subsp. enterica serovars Bovismorbificans, Braenderup, Bredeney, Derby, Hadar, Infantis, Isangi, Kenya, Muenster, and Panama and S. enterica subsp. salamae serovars 41:z10:1,2 and 42:b:e,n,x,z15.
Resistances of E. coli isolates have been reported in a previous paper (Pignato et al., 2009).
WW, wastewater; F, feces.
Antibiotic resistance of isolates from clinical specimens
Out of 144 S. enterica serovar Typhimurium isolates, 24 were sensitive to all tested antibiotics (16.7%), whereas resistance to ampicillin, alone or associated with other resistances, was shown by 102 (70.8%). The resistance pattern AmpChlSulTeStrSp was detected in 32 (22.2%) isolates. Ampicillin resistance alone or associated with other resistances was also detected in Salmonella serovars Blockley, Heidelberg, Goldcoast, and Napoli. Ampicillin resistance was observed neither among the 62 isolates of Salmonella serovar Enteritidis nor among isolates of other less frequent serovars.
Antibiotic resistance of isolates from wastewater
Out of 36 S. enterica serovar Typhimurium isolates, only 1 (2.8%) was sensitive to all tested antibiotics, 35 (97.2%) were resistant to one or more antibiotics, including 33 (91.7%) resistant to ampicillin, and 18 (50%) were resistant to six antibiotics (resistance type AmpChlSulTeStrSp). Ampicillin resistance alone or associated with other resistances was also detected in S. enterica subsp. enterica serovars Blockley and Napoli and S. enterica subsp. salamae serovar 6,7:b:z39. No resistant isolates were found among isolates of other serovars. Antibiotic resistance of 273 E. coli isolates has been reported in a previous study (Pignato et al., 2009).
β-Lactam resistance genes
The bla TEM gene was present in the Salmonella spp. isolates from wastewater and clinical specimens, with the exception of one isolate that did not have any of the bla genes tested and the Salmonella serovar Typhimurium isolates exhibiting resistance patterns AmpChlSulTeStrSp or AmpChlSulTeStrSpNal had the bla PSE-1 gene. The bla TEM gene was also present in all but one ampicillin-resistant E. coli isolates that had the bla SHV gene (Table 2).
Amp, ampicillin; Chl, chloramphenicol; Sul, sulphamethoxazole; Te, tetracycline; Str, streptomycin; Kan, kanamycin; Tmp, trimethoprim; Nal, nalidixic acid; Sp, spectinomycin; ND, none detected; NI, none of the bla genes tested was identified; —, no resistance transfer.
Resistance transfer
Salmonella transconjugants were recovered from two isolates of serovar Blockley and one each of serovars Goldcoast and Heidelberg (Table 2). Among 30 E. coli isolates from wastewater, 13 transferred their resistances as previously reported (Pignato et al., 2009).
Integron detection and characterization
The presence of class 1 integrons, the size of their variables regions, and their gene cassettes are shown in Table 2. Class 1 integrons of 1.0 and 1.2 kb together were detected in all the Salmonella serovar Typhimurium isolates with resistance pattern AmpChlSulTeStrSp from wastewater and clinical specimens. Of the Salmonella spp. isolates from wastewater, only S. enterica subsp. salamae serovar 6,7:b:z39 was positive for a class 1 integron with dfrA1-aadA1 gene cassettes. The same gene cassettes, dfrA1-aadA1, were detected in the Salmonella serovar Goldcoast isolate from clinical specimen. Of the 30 E. coli isolates, 9 (30%) were positive for class 1 integrons with a variable region of 1.0 kb (2 isolates) or 1.5 kb (7 isolates). In all, three distinct integron profiles were identified. The two 1.0-kb integrons contained only one gene cassette, aadA1, whereas the integrons of 1.5 kb contained two gene cassettes: dfrA17-aadA5 (two isolates) or dfrA1-aadA1 (five isolates).
Replicon typing
All the transconjugants were successfully typed by PCR replicon typing, and some isolates proved to carry multireplicon plasmids (Table 2). The majority of the E. coli plasmids (11 out of 13), independently from the carried resistance genes, were positive for the IncF replicon alone or associated with IncFIB (four replicons), IncI1 (three replicons), or IncFIA (two replicons), whereas other plasmids were positive for IncP or IncK replicons. The IncI1 replicon was also detected in Salmonella serovar Goldcoast R plasmid, whereas IncN replicons were detected in the R plasmids of two Salmonella serovar Blockley and one Salmonella serovar Heidelberg isolates sharing similar resistance patterns.
Discussion
Our study was primarily focused on the prevalence of ampicillin resistance in Salmonella isolates from wastewater compared with those from clinical specimens. Quite different rates of ampicillin resistance were observed in the Salmonella isolates according to the serovar. In particular, isolates of the more prevalent serovars, Typhimurium and Enteritidis, differed distinctly for their susceptibility to antibiotics. All the wastewater and clinical Salmonella serovar Enteritidis isolates were susceptible to ampicillin, with low percentages of resistance to other antibiotics, whereas high percentage of ampicillin resistance was found in Salmonella serovar Typhimurium isolates of both origin.
An increase of ampicillin-resistant S. enterica serovar Typhimurium isolates has been observed during the 1990s because of the emergence of an epidemic multidrug-resistant (MDR) strain of definitive phage type (DT) 104. This MDR DT104 clone first emerged in the United Kingdom at the end of the 1980s and has become a major cause of illness in humans and animals in Europe, including Italy, and in the United States (Casin et al., 1999; Glynn et al., 1999; Threlfall, 2000; Carattoli et al., 2002; Mammina et al., 2002; Weill et al., 2006). The multiple antibiotic resistance was due to chromosomal integration of a 43-kb structure called Salmonella genomic island 1 (SGI1), which comprises the bla PSE-1 gene coding for resistance to ampicillin (Boyd et al., 2001). The multidrug resistance region is located at the 3′ end of the SGI1 on a 13-kb region corresponding to a large class 1 integron (Boyd et al., 2001; Levings et al., 2005).
In Italy, MDR Salmonella serovar Typhimurium DT104 is of particular concern, as it represents 20% of all human Salmonella isolates (Cawthorne et al., 2006). Indeed, the Salmonella serovar Typhimurium is the most prevalent serovar accounting for about 40% of all human Salmonella isolates each year and 50% of its isolates belong to the DT104 clone (Cawthorne et al., 2006). In this study, isolates that likely belong to the DT104 clone (similar resistance patterns, presence of the bla PSE-1 gene, and class 1 integrons with two variable regions of 1.0 and 1.2 kb), although phage typing was not performed, represented about 50% (18/36) of the Salmonella serovar Typhimurium isolates from wastewater. These DT104-like isolates were recovered during the 1 year of monitoring from both the treatment plants, indicating the persistent endemic circulation of this clone in the population of the cities conferring their wastewater to the monitored plants. Further, DT104-like isolates were also isolated from clinical specimens although at lower frequency (32/144, 22.2%). Beside this clone, at least another ampicillin-resistant clone characterized by more restricted resistance patterns, presence of the bla TEM gene, and absence of class 1 integrons was endemic in the same population.
Molecular characterization of selected ampicillin-resistant E. coli isolates from wastewater showed that they constituted a rich reservoir of transposable genetic elements. According to the results of other studies (Thomas and Nielsen, 2005; Schluter et al., 2007), it appeared that genetic exchanges could have been occurred among these isolates during their circulation in the intestinal and/or in the wastewater environment, because they shared quite common genetic traits, such as β-lactamase gene type bla TEM and integrons of 1.0 or 1.5 kb with resistance cassettes carrying dfrA1 or dfrA17 genes encoding resistance to trimethoprim and aadA1 or aadA5 genes encoding resistance to streptomycin. The results of PCR replicon typing showed a variety of conjugative R plasmids that shared some more prevalent replicons of the same Inc type. PCR replicon typing has been proved as a sensitive and specific method for identifying phylogenetically related plasmids, opening the possibility of successfully detecting and tracing the diffusion of plasmids and resistance genes (Carattoli et al., 2006). In our experiments, some replicon types were shared by different E. coli and Salmonella isolates showing other common genetic traits. This suggests the possibility that intraspecies and interspecies genetic exchanges have occurred in our environmental and epidemiologic conditions within enteric pathogen and commensal (not only E. coli) microorganisms. Further, considering the variability of the resistance patterns exhibited by the different isolates, gain and/or loss of additional resistance determinants and gene-associated functions have probably occurred across different bacterial clones in humans and/or in the wastewater.
In conclusion, our study indicates that analysis of isolates from wastewater can be a useful and easy epidemiologic tool to monitor the prevalence of antibiotic resistance and genetic elements related to antibiotic resistance in Salmonella clones circulating in the human population.
Footnotes
Disclosure Statement
No competing financial interests exist.