Transfer of plasmid-mediated resistance to tetracycline in pathogenic bacteria from fish and aquaculture environments


  • Editor: Wolfgang Kneifel

Correspondence: Elena Guglielmetti, Centro Ricerche Biotecnologiche, Istituto di Microbiologia, Università Cattolica del Sacro Cuore, Via Milano 24, 26100 Cremona, Italy. Tel.: +39 0372 499 141; fax: +39 0372 499 193; e-mail:


The transferability of a large plasmid that harbors a tetracycline resistance gene tet(S), to fish and human pathogens was assessed using electrotransformation and conjugation. The plasmid, originally isolated from fish intestinal Lactococcus lactis ssp. lactis KYA-7, has potent antagonistic activity against the selected recipients (Lactococcus garvieae and Listeria monocytogenes), preventing conjugation. Therefore the tetracycline resistance determinant was transferred via electroporation to L. garvieae. A transformant clone was used as the donor in conjugation experiments with three different L. monocytogenes strains. To our knowledge, this is the first study showing the transfer of an antibiotic resistance plasmid from fish-associated lactic bacteria to L. monocytogenes, even if the donor L. garvieae was not the original host of the tetracycline resistance but experimentally created by electroporation. These results demonstrate that the antibiotic resistance genes in the fish intestinal bacteria have the potential to spread both to fish and human pathogens, posing a risk to aquaculture and consumer safety.


Bacterial disease is a major problem in aquaculture. Infections of fish by pathogenic microorganisms in the aquatic environment are a serious source of mortality (Toranzo et al., 2005). Several bacterial diseases may be prevented by vaccination. In addition, during disease outbreaks, fish populations may be treated by the removal of infected or dead fish, by chemotherapy, and by the use of antibiotics. Antibiotics have been used, historically, to effectively eliminate bacteria before or during a disease outbreak. Sometimes antibiotics are added in small prophylactic quantities to the feed to promote growth of fish. However, antibiotic residues in the tissues and the increasing incidence of drug-resistant fish pathogens, have reduced the usefulness of antibacterial therapy (Aoki & Takahashi, 1997; Hansen & Olafsen, 1999; Chopra & Robert, 2001).

Tetracycline is the most popular drug in aquaculture due to its efficacy, low cost, and lack of toxicity (Shao, 2001). However, its indiscriminate use has led to a high prevalence of tetracycline resistance genes in marine aquaculture sites (Kim et al., 2004). Similarly, the normal microbiota in fish and in the aquatic environment may be important reservoirs for antibiotic resistance genes (Zhao & Aoki, 1992). Pathogenic bacteria have developed a number of strategies to resist the action of antibiotics, including modification and inactivation of the drug, exclusion of the antibiotic, and alteration of the target. In Gram-positive bacteria, tetracycline resistance genes are carried on a variety of mobile genetic elements, including transposons and self-transferable plasmids (Grohmann et al., 2003; Bertrand et al., 2005). Resistance to tetracycline occurs via two primary mechanisms: an energy-dependent efflux of tetracycline, mediated by an integral membrane protein (Levy, 1992), and protection of ribosomes by a soluble protein, termed ribosomal protection protein (RPP) (Manavathu et al., 1990; Burdett, 1991). The occurrence of the gene that encoded the efflux pump has been demonstrated in various aquaculture environments (Rhodes et al., 2000; Furushita et al., 2003). The only reports of the RPP-encoding genes in the aquatic environment are those by Petersen & Dalsgaard (2003) and Kim et al. (2004), who detected tet(M) in Enterococcus spp. and tet(S) in Lactococcus garvieae isolated from fish.

The widespread distribution of the tet(M)-like genes, the most frequently encountered tet genes in both Gram-positive and Gram-negative pathogens in normal human and animal microbiota, food, and environments, support the hypothesis that the tet genes are exchanged by bacteria from many different source (Roberts, 1996; Chopra & Robert, 2001).

Lactococcus garvieae (previously known as Enterococcus seriolicida) is the etiological agent of Lactococcus infection, which has been particularly devastating in the freshwater culture of salmonid fish and marine-cultured species. Such infections cause significant economic losses when water temperature rises over 16 °C during summer months (Vendrell et al., 2006). Human infection by L. garvieae is rare (Elliot & Facklam, 1996), but another species found in different seafood products, and commonly pathogenic for humans, is Listeria monocytogenes. It can cause serious invasive diseases, primarily in immunologically compromised hosts, the elderly, and neonates (Gellin & Broome, 1989). Listeriosis in adults usually occurs after consumption of contaminated food (WHO Working Group, 1991). Ericsson et al. (1997) reported cases in which the infection outbreak was connected with the consumption of cold-smoked, ‘gravad’ rainbow trout (Oncorhynchus mykiss) and salmon (Salmo salar).

It is therefore relevant to assess the potential transfer of the antibiotic resistance gene among, commensal and pathogenic bacteria that inhabit the gut of fish, in order to better understand the risk associated with the genetic resistance reservoir in this peculiar ecological niche. Here we report experiments that transfer of Lactococcus lactis tet(S) gene from L. garvieae to L. monocytogenes. Donor of this conjugation experiment was made resistant by electroporation of L. lactis plasmid encoding a Tet protein.

Materials and methods

Bacterial strains, media, and antibiotic susceptibility testing

Bacteria included in this study are listed in Table 1. Lactococcus strains were cultivated at 30 °C in M17 medium (Oxoid, Hampshire, UK) supplemented with glucose (GM17). Listeria strains were grown at 37 °C in brain heart infusion (BHI) medium (Difco, Lancashire, UK).

Table 1.   Bacterial strains and plasmids
Strains and/or plasmidsRelevant characteristicsRemarks and/or references
  1. Tet±, tetracycline resistance; Lac±, competence for lactose metabolism; Bac±, competence for production; DSM, Deutsche Sammlung von Mikroorganismen (Braunschweig, Germany); ATCC, American Type Culture Collection, MD, USA.

Lactococcus lactis ssp. lactis KYA-7Plasmid-located tet(S) geneIntestinal sample from salmonid fish (Oncorhynchus mykiss); this study
L. lactis ssp. garvieae DSM 6783TetRecipient for electroporation
L. lactis ssp. garvieae DSM 6783 TF-aTet+TF from DSM 6783 × plasmid tet(S) gene; donor for conjugation; this study
L. lactis ssp. cremoris DSM 4645Lac+, Bac+Used as plasmid size reference (Neve et al., 1984)
Listeria monocytogenes ATCC 7644TetRecipient for conjugation
L. monocytogenes LMK8TetRecipient for conjugation; water in aquaculture; this study
L. monocytogenes LMK16TetRecipient for conjugation; water in aquaculture; this study
Plasmid tet(S) geneTet+Resident plasmid of KYA-7; donor plasmid DNA for electroporation; this study

A broth microdilution method in microtiter plates was used to measure the minimal inhibitory concentrations (MICs) according to CLSI recommendations (2006). Two optimal media were used for the tetracycline microdilution assay: LSM broth (Klare et al., 2005), specially designed for detecting antibiotic susceptibilities in lactic acid bacteria (LAB), for testing Lactococcus spp., and cation-adjusted Mueller–Hinton (CAMHB; Difco) for L. monocytogenes. Tetracycline was purchased from Sigma Aldrich (Steinheim, Germany).

Individual colonies from fresh agar plate cultures were suspended in 5 mL of 0.9% saline to an OD of 0.5 McFarland (corresponds to c. 108 CFU mL−1) and then diluted 1 : 10 in saline. An inoculum of 5 μL (c. 105 CFU mL−1) was added to each well of microtiter plates containing 100 μL of medium. The final concentrations of tetracycline in the medium ranged from 0.25 to 256 μg mL−1, in twofold series. The inoculated plates were subsequently incubated aerobically at 30 or 37 °C for 24 h, depending on the bacterial species.

Detection of antagonistic activity

The previously described well diffusion assay (Moreno et al., 1999) was used for the detection of the antagonistic activity of L. lactis KYA-7. Approximately, 105 CFU mL−1 stationary-phase cells of indicator strains (L. garvieae DSM 6783, L. monocytogenes LMK8, LMK16 and ATCC 7644), were mixed with 20 mL of agar medium, poured in sterile Petri dishes, and allowed to harden. A 5-mm diameter well was made in the solid substrate. The base of each well was sealed with 50 μL of soft agar medium (0.75%), and then filled with an aliquot (50 μL) of overnight culture of L. lactis KYA-7. Detection of antagonistic activity between L. lactis KYA-7 and L. garvieae DSM 6783 were performed on GM17 agar plate. Detection of antagonistic activity between L. lactis KYA-7 and L. monocytogenes strains was performed on BHI agar plates. Plates were incubated overnight at the optimum temperature of the indicator strains. When inhibition zones were present, their diameters were measured.

DNA manipulation

Total genomic DNA from Lactococcus and Listeria strains was extracted using the NucleoSpin® Tissue kit (Macherey-Nagel, Germany), according to the manufacturer's recommendations. Plasmid DNA was isolated using the Wizard® Plus SV Minipreps DNA Purification System kit (Promega, Madison), according to the manufacturer's protocol. We carried out the lysozyme step described by Anderson & McKay (1983). As a probe for Southern hybridization, we used the PCR product (335 bp) obtained with the newly designed primers tetS-FW (5′-GGAGTACAGTCACAAACTCG-3′) and tetS-RW (5′-GGATATAAGGAGCAACTTTG-3′). Labelling and detection were carried out using the DIG DNA labelling and detection kit (Roche, Germany).

Plasmid transformation of Lactococcus species

Lactococcus garvieae DSM 6783 was transformed by electroporation, as described by Holo & Nes (1989). Briefly, to obtain electro-competent cells, a fresh inoculum of overnight-grown culture was made in SGM17 (GM17 containing 0.5 M sucrose), supplemented with 1% glycine until an exponential phase was reached. Cells from 1 mL of exponential-phase culture were washed twice with ice-cold electroporation buffer (0.5 M sucrose, 10% glycerol) and suspended in 200 μL of the same buffer. An aliquot (40 μL) of this suspension was mixed with purified plasmid DNA (0.01–2 μg) and exposed to a single electrical pulse. The pulse was delivered by a Gene-Pulser (Bio-Rad Laboratories, CA), set at 25 μF and 0.2 kV, with a 200-Ω resistor (pulsed controller; Bio-Rad). After incubation for 4 h at 30 °C, the suspension (100 μL) was spread on selective SGM17 plates containing 5 μg mL−1 of tetracycline. Transformants (TFs) were counted after 72 h of incubation at 30 °C. All TF clones were checked at species level by PCR, using the species-specific primers described by Pu et al. (2002). We chose the clone L. garvieae DSM 6783 TF-a, obtained by electroporation, as the donor for conjugation assays.

Mating procedure

The ability of the L. garvieae TF-a to transfer the tet(S) resistance gene was examined by filter mating. Two strains, isolated from water in aquaculture, L. monocytogenes LMK8 and LMK16 and a type ATCC 7644 strain were used as recipients. Before the mating experiment, the ability of the donor strain to grow on PALCAM Listeria selective agar (Difco) plates was checked.

Conjugal transfer was performed on the sterile membrane 0.45-μm filter. Bacterial cells, grown overnight, were mixed with a 1 : 10 ratio (donor : recipient) and collected on a filter by filtration. The filter was washed with 2 mL of 0.9% saline and incubated topside up on BHI agar plate with 0.2 μg mL−1 of tetracycline.

After overnight incubation at 37 °C in mating plates, bacteria were removed from the filter by vigorous shaking in 1.5 mL of saline buffer (Neve et al., 1984). Serial dilutions were spread on PALCAM Listeria selective agar supplemented with 10 μg mL−1 of tetracycline (Perreten et al., 1997).

Conjugation frequency was expressed as the number of transconjugants (TCs) per recipient (CFU) after mating. Colonies were counted after incubation at 37 °C for 72 h. To be sure that the clones obtained by this procedure belonged to the correct species, TCs were identified by API®Listeria test (Biomerieux, Mercy l'Etoile, France).

Detection of tet(S) gene by PCR

To check for the presence of the tetracycline gene in TC and TF clones, PCR assay were performed with tetS-FW and tetS-RW primers described above. The PCR mixture (total volume of 25 μL) contained 1 μL of template DNA preparation, 2 mM MgCl2, 0.2 mM each of dATP, dGTP, dCTP, and dTTP, 10 mM of each primer (Oligomer, Helsinki, Finland), 0.5 U of Taq DNA polymerase, and 1 × PCR buffer. The PCR buffer, nucleotides and polymerase were obtained from Life Science (Fermentas, Finland).

All PCR amplifications were performed in a T3 Thermocycler (Whatman Biometra®, Germany) using the following temperature program: initial denaturation at 95 °C for 3 min, followed with 34 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s, and a final extension step at 72 °C for 10 min. PCR products were separated by electrophoresis on a 1% agarose gel, stained with ethidium bromide and visualized under UV light.

Results and discussion

Screening for antagonistic activity and transformation of Lactococcus by electroporation

In the direct mating experiments, the cell densities of L. garvieae DSM 6783 and L. monocytogenes strains grown with L. lactis KYA-7 were reduced c. 100-fold, compared with the control cultures without L. lactis KYA-7 (data not shown), suggesting antagonistic activities. Inhibitory activity of L. lactis KYA-/was confirmed using against four indicator strains in well experiments. For strains of L. monocytogenes LMK8 and ATCC 7644, the diameter of inhibition zone was 20 mm, while for L. monocytogenes LMK16 and L. garvieae DSM 6783, the diameters were 8 and 5 mm, respectively.

The ability to produce antagonistic substances such as bacteriocins, organic acids, hydrogen peroxide, acetoin, and carbon dioxide is widely distributed among LAB, having an important role in inhibiting the growth of spoilage and pathogenic organisms in food and natural microbiota (Abriouel et al., 2005; Şimşek et al., 2006).

The antagonistic activity of nonpathogenic bacteria could be useful for the inactivation of L. garvieae, a common fish pathogen. However, the effect of this antagonism in vivo, in the fish, remains unknown. The antagonistic effect of L. lactis KYA-7 against L. garvieae DSM 6783 and L. monocytogenes strains LMK8, LMK16, and ATCC 7644 could be a limitation during conjugation in vitro. For these reasons, electroporation was used to obtain a donor strain harboring the plasmid tet(S) gene for further conjugation experiments. Purified large plasmid DNA from L. lactis KYA-7 was used to transform L. garvieae DSM 6783. The results indicated that L. garvieae DSM 6783 TFs were obtained at a low frequency of 3.5 × 103 TFs per microgram of plasmid DNA.

Plasmid extraction of the L. garvieae DSM 6783 TR-a confirmed the presence of a novel plasmid band. This band, at c. 30 kbp, was not present in the recipient strain L. garvieae DSM 6783, but matched the size of a band in the plasmid profile of the donor strain L. lactis KYA-7. In Southern hybridization with a tet(S) probe, the plasmids of donor KYA-7 and DSM 6783 TF-a strains showed positive signals (Fig. 1).

Figure 1.

 Analysis of plasmids isolated from Lactococcus lactis ssp. lactis KYA-7, Lactococcus garvieae DSM 6783 strains before and after tet(S) plasmid transformation by (a) agarose gel electrophoresis and (b) Southern hybridization analysis with tet(S) probe. Lanes: M, DNA Molecular Weight Marker II, DIG-labeled (Roche); 1, supercoiled DNA ladder (Promega); 2, plasmid DNA of L. lactis ssp. cremoris DSM 4645 as plasmid DNA size marker (the size of the indicated plasmids in kbp are 38.8, 27.6, 17.2, 7.4, 5.8, 2.8, and 2.0); 3, L. lactis ssp. lactis KYA-7; 4, L. garvieae DSM 6783 before electroporation; 5, L. garvieae DSM 6783 TF-a after electroporation.

The presence of this plasmid was also associated with an increase of the MIC for tetracycline (0.25 μg mL−1 before the transformation, 256 μg mL−1 in TFs).

Transfer of the tet(S) gene to L. monocytogenes

Although transformation, transduction, and conjugation are practical laboratory methods for transferring of antibiotic resistance, only conjugative transfer appears to be significant in vivo between species and genera (Lacey, 1984; van Elsas, 1992).

In vitro transferability of the tet(S) resistance gene from L. garvieae DSM 6783 TR-a to three strains of L. monocytogenes was investigated using a filter mating procedure. All recipients were susceptible to tetracycline with an MIC of 1 μg mL−1. Frequencies of conjugation and MICs of the TC clones obtained are summarized in Table 2.

Table 2.   Conjugal transfer of tetracycline resistance from Lactococcus garvieae to Listeria monocytogenes
DonorRecipients*Transfer frequency
with Tet (0.2 μg mL−1)
TC per recipient
Tetracycline MIC
(μg mL−1) for tet(S)
positive TCs
  • *

    All recipients with tetracycline MIC of 1 μg mL−1.

L. garvieae DSM 6783 TF-aL. monocytogenes LMK87 × 10−764
L. monocytogenes LMK16<10−10
L. monocytogenes ATCC 76444 × 10−7128

Of the three L. monocytogenes recipient strains, LMK8 and ATCC 7644 were able to receive resistance tet(S) genes from the Lactococcus strain at a frequency of 10−7 TCs per recipient. This occurred only in the presence of a subinhibitory concentration of tetracycline in the mating medium. No TCs were observed with the recipient LMK16. This finding is in agreement with other studies of the transfer of tetracycline determinants in Listeria spp., in which the presence of subinhibitory concentrations of tetracycline in the mating medium favoured both intra- and intergeneric dissemination of tet genes (Doucet-Populaire et al., 1991; Facinelli et al., 1993). Transfer of resistance from Lactococcus to Listeria might occur in the aquatic environment and in the fish where the species cohabit, and where subinhibitory levels of tetracycline may occur, due to use of medicated feed.

The donor strain DSM 6783 TR-a showed an MIC of 256 μg mL−1, while TCs obtained by conjugation using L. monocytogenes LMK8 and ATCC 7644 strains as recipients had lower MICs of 64 and 128 μg mL−1, respectively.

The TCs were confirmed to be the same species as the recipient by API®Listeria test. Genotypic characterization of the transferred plasmid was obtained by plasmid profiling (Fig. 2) and PCR reaction, demonstrating that two tested tetracycline-resistant isolates received the tet(S) gene.

Figure 2.

 Plasmid profiles of the donor Lactococcus garvieae DSM 6783 TF-a (lane 3), recipients Listeria monocytogenes LMK8 (lane 4) and L. monocytogenes ATCC 7644 (lane 5), TCs (lanes 6 and 7) obtained from recipients L. monocytogenes LMK8 and ATCC 7644, respectively. Supercoiled DNA ladder (Promega) and plasmid DNA of Lactococcus lactis ssp. cremoris DSM 4645 were used as reference plasmids (lanes 1 and 2). p, Conjugative plasmid; c, chromosomal DNA.

The ability to transfer genetic material via conjugation is widespread among Lactococcus species, and it has been demonstrated in L. lactis ssp. lactis and cremoris (Moreno et al., 1999; Flórez et al., 2008; Lampkowska et al., 2008). Several studies have described the transfer by conjugation of plasmids and transposons carrying tetracycline resistance genes from Enterococcus and Streptococcus to Listeria, and between Listeria species (Vicente et al., 1988; Facinelli et al., 1993; Perreten et al., 1997), but never from Lactococcus to Listeria.

Contamination of fish products with L. monocytogenes presents a risk to consumer health. The main sources of contamination are brook, river, and other runoff water from the environment, which may contain a high content of organic material (Ben Embarek, 1994). The contamination rate of seafood products with the bacterium varies from zero to >50% (Jinneman et al., 1999). It is extremely difficult to produce fish products that are totally Listeria-free. In Finland, a listeriosis outbreak was connected with the consumption of cold-smoked rainbow trout (Miettinen & Wirtanen, 2005). The Commission Regulation (EC) No. 2073/2005 establishes microbiological criteria regarding acceptable amounts of L. monocytogenes (<100 CFU g−1) in different food products at the time of consumption.

Human listeriosis is a public health problem with a low incidence but high mortality. Because it can be food borne, antibiotic resistance and efficient empirical treatment of infections should be assessed. Miettinen & Wirtanen (2005) have observed that the persistent survival of the L. monocytogenes strain in an environment that is rich in tetracycline, as is water in aquaculture, selects a tetracycline-resistant cluster, causing a difficulty in treating infectious disease.


The use of antibiotics for prophylaxis and animal growth promotion is associated with the risk of selecting for resistance. This makes the transfer of resistance determinants likely in vivo.

In this report, we demonstrated that a large plasmid from L. lactis, which carries tet resistance, can be established in L. garvieae. This species is a common cause of fish Lactococcus infection, and an emerging pathogen affecting a variety of fish species worldwide. It has also been identified in meat products (Rantsiou et al., 2005), in raw cow's milk (Villani et al., 2001), and in humans, in several cases (Fefer et al., 1998).

To our knowledge, this is the first study to demonstrate, by conjugation in vitro, from the fish pathogen L. garvieae to the human pathogen L. monocytogenes of the antibiotic resistance gene tet(S).

In general, the transfer of antibiotic resistance genes is possible within a natural environment (Morelli et al., 1988; Gruzza et al., 1994; Moubareck et al., 2003; Hart et al., 2006). However, the experiments are technically difficult because of the unknown numbers and strains of bacteria that may be present, as well as the difficulties in designing adequate control experiments.

The constructs designed in this study may be useful in controlled in vivo experiments for assessing the likelihood of this kind of transfer. Further experiments are required to characterize the tet(S) plasmid and the in vivo transferability of the resistance.


We are grateful to Angela H.A.M. van Hoek (Rikilt-Institute of Food Safety, Department of Microbiology and Novel Foods, Wageningen, the Netherlands) for designing primers tetS-FW and tetS-RW. We thank Carme Plumbed-Ferrer and Maria Luisa Callegari for supporting the work. Furthermore, we are grateful for the technical assistance provided by Mirja Rekola and Elvi Mäkirinne.