In this study we assessed the occurrence, diversity and conjugative potential of plasmids in integron-carrying Aeromonas and Enterobacteriaceae from wastewaters. Sixty-six strains were included as donors in mating assays using rifampicin-resistant Escherichia coli and Pseudomonas putida recipient strains. The diversity of plasmids from donors and transconjugants (resistant to tetracycline or streptomycin) was evaluated by restriction analysis and replicon typing targeting 19 incompatibility groups. Restriction patterns revealed a diverse plasmid pool present in these strains. Plasmids were assigned to FrepB (Aeromonas salmonicida, Aeromonas veronii, Aeromonas sp., E. coli, Enterobacter sp.), FIC (A. salmonicida, Aeromonas sp.), FIA (Shigella sp.), I1 (A. veronii, Aeromonas sp., E. coli), HI1 (E. coli) and U (Aeromonas media) replicons. Nevertheless, 50% of the plasmids could not be assigned to any replicon type. Among integron-positive transconjugants, FrepB, I1 and HI1 replicons were detected. Results showed that wastewaters enclose a rich plasmid pool associated with integron-carrying bacteria, capable of conjugating to different bacterial hosts. Moreover, replicons detected in this study in Aeromonas strains expand our current knowledge of plasmid diversity in this genus.
horizontal gene transfer
mobile genetic element
Identification and classification of plasmids has been an important issue in recent decades to trace plasmid evolutionary origins and to elucidate their role in environmental processes and microbial adaptation (Johnson & Nolan, 2009a). Classification is usually based on genetic traits related to plasmid maintenance and replication control. Plasmids that use the same replication system belong to the same incompatibility group and compete for stable maintenance. Therefore, plasmids belonging to the same incompatibility group cannot stably coexist in the same cell, although their accessory genes may be different (Couturier et al., 1988).
The importance of plasmids in bacterial adaptation has been reported in several environments, such as soil (Lee et al., 2006), rivers (Shintani et al., 2008) and wastewaters (Verma et al., 2002). Despite the energetic burden, plasmids provide a fitness advantage to their hosts which allow them to persist across bacterial generations (Dionísio et al., 2005). The genetic traits harboured on plasmids may include genes involved in mechanisms such as resistance, energy metabolism, virulence, pathogenicity, symbiosis and/or dissemination, favouring the survival of bacterial hosts under selective pressures (Dionísio et al., 2002).
Conjugation is considered a major pathway for horizontal gene transfer (HGT) among bacteria (Sørensen et al., 2005). It involves direct cell-to-cell contact and DNA exchange usually mediated by a conjugative plasmid. Conjugative plasmids can be highly promiscuous and transfer may occur between different genera or even domains (Ochman et al., 2000). Antibiotic resistance plasmids are found in several bacterial genera, both Gram-negative and Gram-positive. Because of their wide distribution and because they may confer multiple resistance phenotypes, resistance plasmids are of both clinical and environmental concern. Several plasmid families carrying multiple antibiotic resistance determinants have been reported in Aeromonas spp. (Sørum et al., 2003; Picão et al., 2008; Fricke et al., 2009) and Enterobacteriaceae isolates (Carattoli, 2009).
Integrons are genetic structures that enable bacteria to acquire and express gene cassettes, many of them involved in antibiotic resistance (Cambray et al., 2010). Although integrons are transposition defective, they can be mobilized in association with functional transposons and/or conjugative plasmids (Cambray et al., 2010). Despite their relevance in HGT processes, the association of integrons with conjugative plasmids has been poorly addressed in aquatic environments.
Wastewater treatment plants (WWTPs) are important reservoirs of resistance determinants and favourable places for HGT, due to high microbial abundance, high nutrient concentrations and intense selective pressures imposed by antibiotics, detergents and other pollutants (Schlüter et al., 2007). Moreover, it has been shown that natural conjugative plasmids may induce the development of biofilms, which might also increase the chances of cell-to-cell contact and the occurrence of HGT events (Ghigo, 2001). As a result, WWTPs may favour the persistence of plasmids through the treatment process, contributing to the dissemination of integrons and undesirable genetic traits, such as those coding for antibiotic resistance and virulence determinants, to natural waters, soils and eventually the food chain.
Previously, the presence and distribution of integron-carrying bacteria was investigated at different stages of the treatment process in two WWTPs, one treating urban discharges and the other treating wastewaters from a slaughterhouse (Moura et al., 2007, 2012). The present study was performed to investigate the diversity of plasmids in integron-positive strains retrieved from wastewaters, providing data pertaining to the contribution of these environments to the spread of integrons and antibiotic resistance determinants through HGT.
Materials and methods
Bacterial strains and mating assays
Sixty-six integron-positive (intI+) strains belonging to Aeromonas sp. (n = 48) and Enterobacteriaceae (n = 18) previously isolated from urban and slaughterhouse wastewaters (Moura et al., 2007, 2012) were included as donors in mating assays using rifampicin- and kanamycin-resistant Escherichia coli CV601-GFP and Pseudomonas putida KT2442-GFP as recipient strains (Smalla et al., 2006).
Liquid cultures of donor and recipient strains were prepared separately in 10 mL Luria–Bertani broth (LB) and grown overnight with gentle shaking at 28 °C. Recipient and donor strains were mixed (ratio 1 : 1) and centrifuged for 5 min at 6700 g to precipitate cells. Supernatants were discarded and replaced by 1 mL fresh LB. Mixtures were incubated overnight at 28 °C without shaking. Cells were then precipitated by centrifugation (5 min, 6700 g) and washed in 0.9% NaCl solution. Serial dilutions were prepared in 0.9% NaCl and aliquots of 100 µL were spread on Plate Count Agar plates supplemented with rifampicin (50 mg L−1) and streptomycin (50 mg L−1) or with rifampicin (50 mg L−1) and tetracycline (50 mg L−1). Putative transconjugants were grown at 28 °C for 48 h. Assays were run in duplicate. Donor and recipient were also placed on the selective plates for mutant detection.
Putative transconjugants were confirmed by BOX-PCR typing. Profiles were generated by PCR amplification in 25 µL reaction mixtures containing 3.75 mm MgCl2, 0.2 mm dNTPs, 1× Stoffel buffer, 0.2 µm of primer BOX-AIR (5′-CTACGGCAAGGCGACGCTGACG-3′; Versalovic et al., 1991), 2.5 U Stoffel Taq polymerase (Applied Biosystems) and 1 µL of cell suspension prepared in 100 µL of distilled water (∼ 1.0 McFarland turbidity standard).
Amplification was carried out as follows: initial denaturation for 7 min at 94 °C, then 35 cycles of denaturation at 94 °C for 7 min, followed by annealing at 53 °C for 1 min and extension at 65 °C for 8 min, and a final extension at 65 °C for 16 min. Generated profiles were separated in 1.5% agarose gels in 0.5× TBE buffer (50 mm Tris, 50 mm boric acid, 0.5 mm EDTA), at 50 V for 95 min, and stained with ethidium bromide.
Plasmid isolation and restriction analysis
Plasmid DNA from donors and transconjugants was purified using Qiagen Plasmid Mini-kit (Qiagen GmbH, Germany). Diversity of plasmids was evaluated by plasmid restriction analysis using 5 U of PstI (CTGCA↓G) and 5 U of Bst1770I (GTA↓TAC), according to the manufacturer's instructions (Fermentas, Lithuania).
Restriction patterns were visualized in 0.8% agarose gels. Electrophoresis was run at 40 V for 3 h in 0.5× TBE buffer and stained using ethidium bromide. Restriction patterns were compared using GelCompar II software (Applied Maths, SintMartens-Latem, Belgium).
Replicon typing and Southern hybridization with replicon probes
Detection of IncP-1, IncQ, IncN and IncW replicons and integrase genes was performed as previously described (Götz et al., 1996; Moura et al., 2010). Briefly, gels were transferred onto nylon membranes (Hybond-N, Amersham, Germany) and hybridized in middle stringency conditions with PCR-derived specific digoxigenin-labelled probes for intI1, intI2, IncP-1 (trfA), IncQ (oriV), IncN (rep) and IncW (oriV) (Moura et al., 2010).
Detection of IncA/C, IncB/O, IncF (FIA, FIB, FIC, FIIA, FrepB subgroups), IncHI1, IncHI2, IncI1-Iγ, IncK, IncL/M, IncU, IncT and IncY replicons was performed by PCR, using primers and conditions previously described (Carattoli et al., 2005). Results were confirmed by sequencing, except for IncFrep replicons, which were confirmed by Southern hybridization with digoxigenin-labelled probes generated by PCR from positive controls (Carattoli et al., 2005).
Results and discussion
The aim of this study was to evaluate the occurrence, diversity and conjugative potential of plasmids in integron-carrying bacteria from wastewater environments. The presence of plasmid DNA was confirmed in 77% (51 out of 66) of the strains. In the remaining 15 strains (∼ 23%), no plasmids were detected by the plasmid extraction method used. Thus, most of the strains analysed harboured at least one plasmid, these strains being retrieved from all stages of the treatment process, including from final effluents (Table 1). Nevertheless, the presence of additional plasmids cannot be excluded. Although the extraction methodology is appropriate for purification of plasmids of low and high copy number, it is limited to plasmids up to 150 kb, according to manufacturer's data. In addition, the presence of very low-copy plasmids (fewer than five copies) may be constrained by this approach.
Characterization of donor strains used in this study in terms of phylogeny, presence at WWTPs, antibiotic resistance profile, plasmid incompatibility group and generation of transconjugants in mating assays
↵Incompatibility grouping was determined by replicon typing and Southern hybridization (n.d., not detected).
↵Ec (E. coli CV601-GFP) and Pp (Pseudomonas putida KT2422-GFP) refer to the recipient strains; TET (tetracycline) and STR (streptomycin) refer to the selective markers used; –, no transconjugants observed.
Plasmid restriction analysis showed a diverse plasmid pool in intI+ strains from wastewaters. In total, 45 different plasmid restriction patterns (similarity < 98%) were obtained (Fig. 1). No restriction patterns were recovered from six strains (MM.1.10, MM.1.11, MM.1.12, MM.1.14, MM.1.15, MM.1.26), due to low plasmid DNA concentration, which may be due to lower plasmid DNA extraction efficiency and/or very low-copy plasmid number. Restriction patterns did not cluster by species, type of effluent or treatment stage, suggesting a high diversity of backbones and/or accessory elements present in these strains. The results reinforce that wastewaters are reservoirs of diverse mobile genetic elements and hotspots for HGT, as previously reported (Schlüter et al., 2007; Moura et al., 2010).
Unweighted pair group method with arithmetic mean (UPGMA) dendrogram based on a Dice similarity correlation generated from the plasmid restriction patterns of intI+ donor strains obtained from urban (n = 24) and slaughterhouse (n = 21) wastewaters. Dispersion at the WWTP is also indicated: RW, raw waters; PD, primary decantation tank; HT, homogenization tank; AT, aeration tank; SR, sludge recirculation; FE, final effluent.
Among donor strains, plasmids were assigned to FrepB (Aeromonas salmonicida, Aeromonas veronii, Aeromonas sp., E. coli, Enterobacter sp.), FIC (A. salmonicida, Aeromonas sp.), FIA (Shigella sp.), I1 (A. veronii, Aeromonas sp., E. coli), HI1 (E. coli) and U (Aeromonas media) replicons (Table 1 and Fig. 1). Although the presence of broad-host-range IncN, IncQ, IncW and IncP-1 plasmids had been detected in total community DNA obtained from the same environments (Moura et al., 2010), none of the donor strains gave positive hybridization signals using probes targeting these groups. Other studies dealing with total community DNA and exogenous isolation of plasmids from urban wastewaters also suggested that broad-host-range plasmids, in particular those belonging to the IncP-1 group, are abundant in wastewater environments (Dröge et al., 2000; Heuer et al., 2002; Schlüter et al., 2007; Bahl et al., 2009). Thus, results obtained here suggest that the hosts of broad-host-range plasmids may probably be noncultivable bacteria and/or bacteria from other taxa than those focused in this study.
To date, reported replicons in Aeromonas spp. have been limited to IncU and IncA/C, identified in different aquatic environments. IncU replicons have been reported in Aeromonas caviae, A. media, Aeromonas allosaccharophila, Aeromonas hydrophila and A. salmonicida strains isolated from rivers (Cattoir et al., 2008), lakes (Picão et al., 2008), fish farms and hospital sewage (Rhodes et al., 2000), often associated with tetracycline and/or quinolone resistance determinants. IncA/C plasmids have been reported in A. hydrophila and A. veronii strains isolated from fish carriage water (Verner-Jeffreys et al., 2009). Therefore, the types of replicons found in this study in Aeromonas strains (FrepB, FIC, I1 and U) expand our current knowledge on plasmid diversity in this genus, increasing the range of molecules capable of replication in Aeromonas spp. IncF and IncI1 type plasmids have been frequently reported worldwide in clinical Enterobacteriaceae, associated with the spread of resistance genes towards extended-spectrum beta-lactams, aminoglycosides and quinolones (Carattoli, 2009). In addition, the presence of IncI1 replicons has also been reported in bacteria isolated from domestic and wild animals, as well as food products (Carattoli, 2009). The presence of IncF-type and IncI1 plasmids in Aeromonas strains again highlights the importance of members of this genus as hosts of mobile genetic elements (Rhodes et al., 2000; Sørum et al., 2003; Moura et al., 2007, 2012; Cattoir et al., 2008; Picão et al., 2008; Verner-Jeffreys et al., 2009; Kadlec et al., 2011). In addition, the common association of F-type replicons to virulence traits, such as colonization factors and toxins in E. coli (Johnson & Nolan, 2009b), as well as their presence in treated effluents, raises concern regarding the possible dissemination of such traits to natural environments, agriculture fields and the food chain.
Despite the diversity of replicons found among donor strains, 50% of plasmids remained unknown, possibly due to the type of approach used, which relied on the classification of plasmids belonging to classic Inc groups, thus failing to identify novel or divergent replicons (Carattoli, 2009).
In total, plasmids from approximately 73% (41 out of 56) of the donor strains with tetracycline and/or streptomycin intermediate or resistance phenotypes transferred successfully to recipient strains under the conditions tested (Table 1). Among Aeromonas spp., plasmids from 70% (28 out of 40) of donor strains transferred successfully to at least one recipient strain, of which 10% (four out of 40) generated transconjugants with both recipient strains. Among Enterobacteriaceae, plasmids from 81.3% (13 out of 16) transferred to at least one recipient strain, of which 18.8% (three out of 16) transferred to both recipient strains. In previous studies, transfer efficiencies ranged between 10−5 and 10−6 transconjugants per recipient cell for these Aeromonas donors, whereas among Enterobacteriaceae rates were 10−5 transconjugants per recipient cell (Moura et al., 2007, 2012). Although plasmids of narrow host range have difficulty replicating in distantly related hosts, both Aeromonas and Enterobacteriaceae strains from all stages of the treatment process, including final effluent, have generated transconjugants using E. coli and P. putida as recipient strains (Table 1).
Accessory genetic modules, such as integrons, are known to be integrated among functional plasmid backbone modules. Overall, 15% (10 out of 66) of donor strains analysed using this methodology harboured plasmid-borne integrons. A similar prevalence was reported by Tennstedt et al. (2003), who detected the presence of class 1 integrons in 12.4% of resistance plasmids obtained by exogenous isolation from an urban WWTP.
Class 1 integrons have been associated with conjugative plasmids belonging to narrow- (IncFII, IncF, IncC, IncH1, IncHI2, IncU and IncL/M) and broad-host-range incompatibility groups (IncP-1, IncW, IncN, IncA/C) (http://integrall.bio.ua.pt, Moura et al., 2009). Class 2 integrons have been mostly associated with conjugative IncF, IncL/M, IncN and IncP-1α plasmids in E. coli, Klebsiella pneumoniae and Pseudomonas aeruginosa (http://integrall.bio.ua.pt, Moura et al., 2009). In this study, plasmid-borne class 1 integrons were detected in FrepB, FIA, I1 and HI1 in E. coli (Table 1), whereas the replicon type of plasmid-borne class 2 integron in E. coli MM.2.2 could not be assigned. The diversity of restriction patterns obtained from intI+ transconjugants is shown in Fig. 2. Restriction patterns from donors did not cluster with those from intI+ transconjugants (data not shown), suggesting that only a fraction of plasmid population in donor strains was efficiently transferred to or stably replicated in the recipient strains. Also, plasmid transfer could be limited by the selective markers used. The extensive dissemination of plasmid-borne integrons is thought to result from the intensive use of antibiotics and heavy metals in clinical, agricultural and industrial practices, leading to the coselection of class 1 integrons associated with Tn21 transposons that carry the mer operon conferring resistance to mercury (Liebert et al., 1999).
UPGMA dendrogram based on a Dice similarity correlation generated from the plasmid restriction patterns of intI+ transconjugants obtained in this study. Assigned incompatibility groups are also indicated (n.d., not detected); TET (tetracycline) and STR (streptomycin) refer to the selective markers used in mating assays.
In contrast to the results obtained by Moura et al. (2007), no intI+ transconjugants were obtained for strains MM.1.3, MM.2.11 and MM.2.6. This could be due to the use of different methodologies, such as temperature of incubation and additional centrifugation steps, that may affect formation or integrity of pili and plasmid stability (Friehs, 2004). As discussed before, the establishment of a standardized methodology for plasmid transfer analysis would be recommended to allow the systematic testing of conjugative transfers in microbial populations (Sørensen et al., 2005).
In conclusion, these findings expand our current knowledge of plasmid diversity in wastewaters and emphasize the role of these environments in the spread of integrons and antibiotic resistance determinants through HGT. Future work focusing on full sequencing of plasmids which could not be assigned to known groups will allow us to elucidate the diversity of backbones and accessory modules occurring in these environments.
This work was supported by Fundação para a Ciência e a Tecnologia (FCT), through project POCTI/BME/45881/2002 and grants SFRH/BD/19443/2004 and SFRH/BPD/72256/2010 (A.M.), SFRH/BPD/65820/2009 (C.O.) and SFRH/BPD/63487/2009 (I.H.). We thank Ellen Krögerrecklenfort (Julius Kühn-Institut, Germany) for technical assistance and Alessandra Carattoli (Department of Infectious, Parasitic and Immune-Mediated Diseases, Istituto Superiore di Sanità, Italy) for providing replicon typing control strains.
(2000) Distribution of oxytetracycline resistance plasmids between aeromonads in hospital and aquaculture environments: implication of Tn1721 in dissemination of the tetracycline resistance determinant tetA. Appl Environ Microbiol 66: 3883–3890.
(2006) Increased abundance of IncP-1beta plasmids and mercury resistance genes in mercury-polluted river sediments: first discovery of IncP-1beta plasmids with a complex mer transposon as the sole accessory element. Appl Environ Microbiol 72: 7253–7259.