|
Occurrence and fate of antibiotic resistant bacteria in sewage 5. Spread of resistant bacteria and resistance genes by municipal sewage effluentsThe results presented in Chapter 4 showed that, although in low numbers, multiple-resistant bacteria are able to survive sewage treatment and reach natural aquatic habitats through their presence in treated sewage. As some types of multiple-resistant bacteria are unlikely to occur naturally in the aquatic environment, a possible risk could be that novel resistance genes are taken up by the indigenous microflora and spread by mechanisms of horizontal gene transfer. The actual environmental impact incurred depends on the ability of such multiple-resistant bacteria to survive in the aquatic environment, retain their antibiotic resistance properties and transfer resistance genes to the indigenous microflora. In the Part III of the project, we investigated the fate of multiple-resistant bacteria occurring in treated sewage, and more generally, the impact of municipal sewage effluents on the spread of antibiotic resistance. Multiple-resistant strains isolated from treated sewage were studied for their ability to survive and retain antibiotic resistance in natural waters (section 5.1). The transfer of tetracycline resistance genes from bacteria in sewage to bacteria in natural aquatic habitats was investigated by laboratory mating experiments (section 5.2). Finally, the impact of municipal sewage effluents on spread of antibiotic resistance was evaluated by comparing the occurrence of resistant bacteria in blue mussels exposed to treated sewage and blue mussels collected from unpolluted sites (section 5.3). 5.1 Survival in the environment of resistant bacteria originating from sewageThe survival in natural waters of three multiple-resistant strains isolated from the effluent of the Lynetten plant was studied by laboratory seawater microcosms (section 5.1.1) and membrane-filter chambers immersed in a freshwater pond (section 5.1.2). The strains represented three different bacterial species: Acinetobacter johnsonii, Escherichia coli and Citrobacter freundii. Survival experiments were performed both in the presence (i.e. untreated water) and in the absence (i.e. autoclaved water) of the indigenous microflora. A detailed description of the methods used is provided in Chapter 2 (section 2.5). 5.1.1 Survival in laboratory seawater microcosmsGrowth of the multiple-resistant strains was observed during the 48 hours following the inoculation of the strains into the microcosms (Fig. 5.1). The initial growth of the multiple-resistant strains was likely to be due to the incubation conditions used to maintain the microcosms in the laboratory rather than to a particular ability of the strains to multiply in seawater per se, since similar growth was also observed in the microcosm containing the indigenous microflora alone (Fig. 5.1). After 48 hours of incubation, the numbers of multiple-resistant strains gradually declined (Fig. 5.1), whereas, the total numbers of culturable bacteria remained stable at approximately 105 CFU/ml in all microcosms (Fig. 5.1). All three multiple-resistant strains survived longer in autoclaved seawater (Fig. 5.1B) than in untreated seawater (Fig. 5.1A). A similar finding was seen in a previous study investigating the survival of another E. coli strain in seawater from Køge Bugt 66. The reduced survival of E. coli in untreated seawater could be due to both antagonism of the indigenous microflora and predation by protozoa. Figure 5.1. Surprisingly, the environmental species A. johnsonii showed a more rapid die-off compared with the enteric bacteria E. coli and C. freundii. In fact, the numbers of A. johnsonii strain fell below the detection limit (10 CFU/ml) after 14 days of incubation in untreated water, and after 30 days of incubation in autoclaved water, whereas, the E. coli and C. freundii strains were still detected after 30 days (Fig. 5.1). The E. coli strain used in this study survived longer in seawater compared with a previously investigated laboratory strain of E. coli (i.e. E. coli K12), for which a survival of only five days was observed under similar laboratory conditions 66. Physiological and/or structural changes associated with the exposure to various stressful conditions (e.g. antibiotic selective pressure, survival in sewage, survival of sewage treatment, etc.) could enable multiple-resistant bacteria in sewage to survive environmental stresses compared with the laboratory strains generally used in this kind of experiments. This study demonstrated that multiple-resistant bacteria occurring in municipal sewage effluents were able to survive in seawater for at least one month following their inoculation into the microcosms. This result is particularly interesting in consideration of the fact that a low bacterial inoclula was used in comparison with previous studies concerning bacterial survival. The temperature in the laboratory microcosms (26° C to 30° C during the day) was higher compared with natural conditions. However, this should not detract from the validity of the result, as previous studies have demonstrated that E. coli survive longer in seawater at low temperatures 66. 5.1.2 Survival in a freshwater pondThe multiple-resistant strains showed a more rapid die-off in membrane-chambers immersed in a freshwater pond compared with laboratory seawater microcosms. In the chamber containing untreated pond water (Fig. 5.2A), the numbers of the multiple-resistant strains fell below the detection limit (1 CFU/ml) after either 21 days (A. johnsonii and E. coli strains) or 28 days (C. freundii strain). In the chamber containing autoclaved pond water (Fig. 5.2B), E. coli and C. freundii strains survived slightly longer compared with the A. johnsonii strain. This was also the case in the chamber containing untreated pond water. Only the C. freundii strain was recovered after 28 days, although at very low numbers (2 CFU/ml). The numbers of total bacteria were constant in the chamber containing untreated pond water, as well as outside of the chambers. As for the previous experiment conducted in laboratory seawater microcosms (section 5.1), the A. johnsonii strain survived for a shorter period compared with the other two multiple-resistant strains under study. The strain could not be recovered after 28 days of incubation in the pond, even when an enrichment procedure in peptone buffered water was used for detection of damaged and stressed cells. Therefore, it appeared that the A. johnsonii strain was no longer present in the chambers. The use of the enrichment procedure in peptone buffered water revealed that the E. coli strain and C. freundii strains were both present in the two chambers after 28 days. Bacterial isolates (n=15) obtained following inoculation of the enrichment culture on the selective medium for these two strains (i.e. MacConkey agar with added antibiotics) showed the same colony morphology, resistance pattern, plasmid profile and ribotype of the respective C. freundii strain (n=14) and E. coli strain (n=1). These results confirm that the multiple-resistant isolates obtained after 28 days were identical to the test strains initially inoculated into the chambers. After 28 days of incubation in the pond, the E. coli strain and C. freundii strains could be recovered from the first two enrichment dilutions (10-1 and 10-2), but not from further dilutions, indicating that the level of the two strains in the chambers was between 102 and 103 CFU/ml. A proportion of the two strain populations were probably in a stressed state since lower bacterial densities were detected by direct plating on the selective media (Fig. 5.2). Characterisation of the bacterial isolates obtained from the chambers after 28 days of incubation in the pond revealed that the strains generally maintained their original plasmid profiles and multiple resistance properties. Only three isolates showed slight variation in the number of plasmid bands compared with the strain originally inoculated into the chambers. Rare differences were observed with regard to the level of susceptibility to one antibiotic (i.e. cefoxitin), with two isolates showing larger inhibition zone diameters (32 mm) compared with the strain originally inoculated into the chamber (10 mm). Figure 5.2. No bacteria showing the same multiple resistance patterns of the test strains were recovered in the pond. Furthermore, bacteria isolated randomly on MacConkey agar without antibiotics showed phenotypic and genotypic traits different from those of the test strains, indicating that the reduction in the numbers of the multiple-resistant strains observed during the experiment was actually due to bacterial die-off and not to loss of their multiple resistance properties. The results of this experiment showed that two of the three multiple-resistant strains under study were able to survive in the freshwater pond for at least 28 days. Furthermore, the two strains maintained their multiple resistance properties following one month of incubation under natural conditions. Therefore, it appears that multiple-resistant bacteria occurring in municipal sewage effluents can survive in natural freshwater environments for relatively long periods. 5.2 Transfer of resistance genes from sewage to aquatic bacteriaThe possibility that tetracycline resistance is transferred from bacteria in sewage to bacteria in natural aquatic environments was studied under laboratory conditions. Mating experiments were carried out using unrelated tetracycline-resistant Acinetobacter strains isolated from sewage (n=10), aquacultural habitats (n=5) and clinical specimens (n=5) as donors and a tetracycline-sensitive Acinetobacter strain isolated from an unpolluted stream as a recipient (section 5.2.1). Furthermore, tetracycline-resistant Acinetobacter isolates from sewage (n=10), fish farms (n=5) and clinical specimens (n=35) were analysed by PCR for the occurrence of tetracycline resistance genes of the classes Tet A to E, with the aim to determine whether the same genes occur in Acinetobacter populations inhabiting different environments (section 5.2.2). 5.2.1 Laboratory mating experimentsAmong the 20 Acinetobacter strains tested as potential donors of tetracycline resistance, transfer was demonstrated from only three aquatic strains, two from sewage and one from an aquaculture habitat (Paper 3). The two sewage strains capable of transferring tetracycline resistance originated from sewers receiving waste effluent from a hospital (strain LUH 5618) and a pharmaceutical plant (strain LUH 5613) (see Chapter 3). Transfer of tetracycline resistance was not apparent from any of the clinical A. baumannii strains to the aquatic recipient strain used in the laboratory matings. Transfer did not occur when DNA from the donor strains was added to the recipient cultures and was not affected by the presence of deoxyribonuclease I, suggesting a conjugative nature of the transfer. Multiple plasmids of a relatively small size (<36 kb) were transferred from the donor strain LUH 5613 into the recipient strain (Paper 3). In the case of the donor strain LUH 5618, the transfer of tetracycline resistance was apparently not mediated by plasmids, since novel bands were not observed in the plasmid profile of the recipient strain (Paper 3). This laboratory experiment showed that transfer of tetracycline resistance from sewage bacteria to bacteria living in natural aquatic habitats is possible. However, the limited number of strains used in the mating experiments does not permit broad conclusions on the frequency of such a transfer occurring in nature. Additionally, transfer did not occur between distantly related Acinetobacter species, suggesting the existence of physical or physiological barriers limiting the exchange of antibiotic resistance genes between different bacterial species belonging to the same genus. 5.2.2 Distribution of tetracycline resistance genesAmong the 15 aquatic Acinetobacter strains tested, three strains contained Tet B (Paper 3). The remaining aquatic strains contained unspecified tetracycline resistance determinants, which did not belong to any of the common classes occurring in Gram-negative bacteria (Tet A, B, C, D, E, G and M). The three strains containing Tet B had previously been isolated from sewers receiving waste effluent from a pharmaceutical plant (see Chapter 3). The three strains belonged to different species according to both phenotypic and genotypic identification, indicating that Tet B was widespread in the Acinetobacter population of this habitat. A different distribution of tetracycline resistance determinants was observed in clinical strains in comparison with the aquatic strains. Among the 35 clinical strains tested, 33 strains contained either Tet A (n=16) or Tet B (n=17) (Paper 3), indicating that these two classes of tetracycline resistance genes are widely distributed in Acinetobacter populations of hospital environments. The different distribution of tetracycline resistance genes in clinical and aquatic strains indirectly provides evidence that the predominant genes occurring in environmental Acinetobacter populations do not originate from clinical environments. 5.3 Occurrence of resistant bacteria in blue mussels exposed to treated sewageThe occurrence of resistant bacteria was studied in blue mussels collected from sites exposed to treated sewage (i.e. the outlets of the Avedøre and Lynetten plants) and a control site not exposed to treated sewage (i.e. Limnfjorden). Numbers of resistant bacteria were determined for both total culturable bacteria and E. coli as previously described (sections 2.2.5 and 2.2.6). The numbers of resistant bacteria in blue mussels either exposed or not exposed to treated sewage were compared to detect possible associations between antibiotic resistance and exposure to treated sewage. 5.3.1 Antibiotic resistance of total culturable bacteria in blue musselsOccurrence of ampicillin resistance and to a lesser extent nalidixic acid resistance was more frequent than gentamicin and tetracycline resistance in the bacterial flora of mussels exposed to treated sewage (Table 5.1). A different distribution of antibiotic resistance was observed in the bacterial flora of mussels collected from Limfjorden, with nalidixic acid-resistant bacteria being surprisingly more frequent compared with ampicillin-resistant bacteria. Although at very low percentages (# 0.1%), multiple-resistant bacteria occurred at the outlets of the two treatment plants. Multiple-resistant bacteria were also isolated from mussels collected at 100 m from the outlet of the Lynetten plant, but not from mussels collected at 100 m from the outlet of the Avedøre plant. This suggests that the presence of multiple-resistant bacteria in mussels living in the proximity of municipal sewage effluents depends on the specific conditions occurring at each treatment plant. Higher percentages of ampicillin resistance were found in mussels exposed to treated sewage (12.9 to 95.5%) in comparison with mussels not exposed to treated sewage (1.5 to 5.4%). However, interpretation of the data on ampicillin resistance was difficult due to extremely variable percentages of resistant bacteria found in mussels collected from the same sites at different times (Table 5.2). The percentages of gentamicin and tetracycline resistance were low (<3%) independent of the origin of the mussels and the time of sampling. Table 5.1.
A, ampicillin; G, gentamicin; N, nalidixic acid; T, tetracycline; AGT, ampicillin, getamicin and tetracycline; AGNT, ampicillin, getamicin, nalidixic acid and tetracycline; Avedøre 1, outlet; Avedøre 2, ca. 100 m from the outlet; Lynetten 1, outlet; Lynetten 2, ca. 100 m from the outlet; N.D., not detected. At sites exposed to treated sewage, a correlation was generally found between the percentages of antibiotic resistance and the distance from the outlets, with higher percentages of resistant bacteria observed in mussels collected from the outlets compared with mussels collected 100 m from the outlets (Table 5.2.). The only exception was the November sampling at the Lynetten plant where an opposite trend was observed (Table 5.2). This was probably because of the presence of strong currents in the direction of the sampling site situated 100 m from the outlet. Table 5.2.
A, ampicillin; G, gentamicin; N, nalidixic acid; T, tetracycline; AGT, ampicillin, getamicin and tetracycline; AGNT, ampicillin, getamicin, nalidixic acid and tetracycline; Avedøre 1, outlet; Avedøre 2, ca. 100 m from the outlet; Lynetten 1, outlet; Lynetten 2, ca. 100 m from the outlet; N.D., not detected. This study demonstrates that, while resistance to single antibiotics can be found in any environment, including habitats characterised by low levels of pollutants, multiple-resistance to three or four different classes of antibiotics is not likely to occur in natural aquatic bacterial populations. Bacteria demonstrating multiple-resistance to ampicillin, gentamicin and tetracycline were found only in blue mussels exposed to treated sewage, confirming that municipal sewage effluents are likely to represent an important source for the dissemination of these bacteria in the environment. 5.3.2 Antibiotic resistance of E. coli in blue musselsE. coli was only detected in blue mussels exposed to treated sewage, although only sporadically (Table 5.3). This situation did not allow comparison of data on antibiotic resistance between different sampling sites and times. Therefore, E. coli could not be used as a bacterial indicator for monitoring antibiotic resistance in blue mussels. The selective medium used for enumeration of E. coli (i.e. TBX agar) could be usefully employed for microbiological analysis of blue mussels. Among 43 representative bacterial isolates tested by the API 20E identification system, 21 isolates (48.8%) were identified as E. coli with either a very good or good identification score, 14 isolates (32.6%) were identified as E. coli with a low discrimination profile, and 8 isolates (18.6%) showed a doubtful or unacceptable profile. Accordingly, the proportion of verified E. coli on the medium varied from 50% to 82% depending on the interpretation of the identification scores obtained by the API 20E identification system. Table 5.3.
A, ampicillin; G, gentamicin; N, nalidixic acid; T, tetracycline; AGT, ampicillin, getamicin and tetracycline; AGNT, ampicillin, getamicin, nalidixic acid and tetracycline; Avedøre 1, outlet; Avedøre 2, ca. 100 m from the outlet; Lynetten 1, outlet; Lynetten 2, ca. 100 m from the outlet; N.D., not detected. 5.4 ConclusionsThe results of the survival experiments demonstrate that multiple-resistant strains isolated from treated sewage survived and retained antibiotic resistance for the duration of the study (28 days) in both seawater and freshwater. Above all, the multiple resistance properties of the strains under study remained unchanged after one month of incubation under natural conditions. Therefore, it seems that multiple-resistant bacteria occurring in municipal sewage effluents have sufficient time to transfer resistance genes to indigenous aquatic bacteria once they are released into natural aquatic environments. Indeed, transfer of tetracycline resistance was shown to occurr under laboratory conditions from Acinetobacter strains isolated from sewage to a recipient strain originating from an unpolluted freshwater habitat. However, the actual ability of strains originating from sewage to transfer antibiotic resistance genes to aquatic bacteria under natural conditions needs further evaluation through laboratory experiments using larger numbers of recipient and donor strains resistant to different antibiotics, as well as in situ experiments. The lack of transfer of tetracycline resistance from clinical to aquatic Acinetobacter strains and the differences observed in the distribution of tetracycline resistance genes between the two bacterial populations, suggest that most tetracycline-resistant bacteria occurring in sewage and aquacultural habitats do not originate from clinical environments. Bacteria that were multiple-resistant to ampicillin, gentamicin and tetracycline were found in treated sewage and in blue mussels collected at the outlets of municipal sewage effluents, but not in seawater, pond water or blue mussels collected from sites not exposed to treated sewage. This finding substantiates the hypothesis that municipal sewage effluents contribute to the dissemination of multiple-resistant bacteria in the environment. Consequently, future studies investigating the impact of munipal sewage effluents on the spread of antibiotic resistance should focus on the occurrence of multiple-resistant bacteria rather than on the occurrence of bacteria resistant to single antibiotic compounds.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||