Four strains of S. xylosus were used. Three strains isolated from dry fermented sausages are resistant to tetracycline and sensitive to minocycline, their resistance occurring by a mechanism of active efflux encoded by the tetK gene as determined by hybridization of a diagnostic microarray including 10 tet gene probes commonly found in Gram-positive bacteria, i.e., tetA, tetB, tetK, tetL, tetM, tetP, tetS, tetT, tetU, and tetW (Even et al., 2010). The C2a strain is derived from the type strain DSM20267 cured of its endogenous plasmid (Götz et al., 1983). This strain is sensitive to tetracycline and resistant to rifampicin. We selected a spontaneous fusidic acid-resistant mutant from the C2a strain as a recipient strain.

The S. xylosus strains were cultured in Brain Heart Infusion (BHI, Difco) under aerobic conditions (1:10 volume to flask ratio, 150 rpm) or on BHI Agar (Difco) at 37°C. The BHI medium, when required, was supplemented with antibiotics at the following concentrations: 10 μg/ml tetracycline (Tet), 25 μg/ml rifampicin (Rif), and/or 25 μg/ml fusidic acid (Fus) (Sigma-Aldrich).

Transfer experiments were performed by filter mating. The donor and recipient strains were cultivated separately overnight with the appropriate resistance selection to reach about 8 × 10⁸ CFU/ml. Cells were mixed in a donor/recipient ratio of 1:3 or 1:10. Typically, 350 or 100 μl of donor and 700 or 900 μl of recipient, respectively, were pelleted and suspended in 400 μl of BHI and dropped onto 0.45 μm cellulose nitrate membrane filters (Whatman). The filters were incubated on BHI agar plates for 18 h at 37°C. After mating, cells were recovered from the filters by vortexing in 1 ml of BHI. Transconjugants were selected by plating serial dilutions of the mating suspension on BHI agar supplemented with Tet, Rif, and Fus. The plates were incubated for 48 h at 37°C. After incubation, the resulting colonies were picked up and replicated on BHI containing Tet, Rif, and Fus. Likewise, donor and recipient strains were plated as control onto BHI agar containing Tet or Rif and Fus, respectively. The frequency of transfer was expressed as the number of conjugants per recipient. Presumptive transconjugants were confirmed by antibiotic disk diffusion, PCR-based detection of tetK gene, plasmid profiling, and PFGE typing as described below.

Batter was prepared by grinding raw materials (lean pork 80%, pork back fat 20%) to obtain particles of 6 mm diameter and was seasoned with dextrose (0.5%), salt (2.6%), saltpeter (0.03%), and pepper (0.15%). It was inoculated with starter culture Lactobacillus sakei 10⁶ CFU/g. The batter was then divided into two batches that were inoculated by S. xylosus strains XIV 10B1 and C2a at about 10⁶ CFU/g each in batch 1 and 10⁸ CFU/g each in batch 2. The mixtures were stuffed into natural casings. After stuffing, sausages were dipped into a Penicillium surface suspension and hung vertically in a temperature and humidity controlled incubator (ARCOS, France) to carry on the ripening process (fermentation and drying). The products were fermented at 22/24°C for 6 days 96–94% RH (relative humidity), then dried for 24 days at 13°C/14°C and 80–82% RH.

Samples were taken at T0 (after inoculation and stuffing), T6 (end of fermentation stage), and T30 (end of drying). The pH of the samples was recorded using a pH meter MP230 (Mettler Toledo, Viroflay, France) with a pH probe (Inlab 427 penetration probe; Mettler Toledo). The water activity (aw) was measured with an aw-sprint TH500 (Novasina, Roucaire, France). Lactic acid bacteria (LAB) were enumerated on MRS agar (Oxoid) pH 5.7 incubated in anaerobic conditions at 30°C for 72 h. Staphylococci were enumerated on Chapman (Oxoid) after incubation for 24–48 h at 37°C. After counting, a replica-plating procedure was performed on BHI agar supplemented with appropriate antibiotics. Bacterial enumeration was also directly performed on BHI agar supplemented with (1) Rif and Fus, (2) Tet, or (3) Tet, Rif, and Fus. The plates were incubated for 24–48 h at 37°C. Rapid identification of S. xylosus was performed by direct colony PCR using species-specific primers as described (Morot-Bizot et al., 2003; Corbière Morot-Bizot et al., 2004). Presumptive transconjugants were confirmed as described below.

Antibiotic susceptibilities were evaluated by the disk diffusion method according to the guidelines of the CA-SFM (2017) with Mueller Hinton agar (Sigma-Aldrich). The following antimicrobial susceptibility test disks (Bio-Rad) were used: fusidic acid (10 μg), minocycline (30 μg), penicillin G (6 μg), rifampicin (5 μg), and tetracycline (30 μg). Inhibition zones were measured, and the susceptibility was determined as suggested by the CA-SFM standards.

Total DNA from S. xylosus was isolated by the method of Marmur (1961). Approximately 10 ng of total DNA was used as a PCR template. The primers used for tetK amplification were GTAGCGACAATAGGTAATAGT and GTAGTGACAATAAACCTCCTA (Strommenger et al., 2003). PCR was performed in a 50 μL reaction mixture containing 5 pmol of each primer, 200 μM of each dNTP, 1.5 mM MgCl₂, 1× Taq DNA polymerase buffer, and 0.5 U of Taq DNA polymerase (Promega). The amplification cycle was as follows: 3 min at 94°C; 30 cycles of 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C; and a final extension of 4 min at 72°C. The amplification products were analyzed on 2% agarose gel and visualized by ethidium bromide. PCR product of 360-bp was further checked by sequencing.

Plasmid DNA from S. xylosus was extracted using the QIAprep Miniprep Kit (Qiagen) after cell lysis with 16 μg/ml lysostaphin. The plasmid content was verified on a 0.8% agarose gel and visualized by ethidium bromide or transferred onto positively charged nylon membranes (Hybond-N+, Amersham Biosciences). Southern blot hybridization was performed using the tetK PCR product (360 bp long) from S. xylosus XIV 10B1 as a probe. The PCR product was first purified with a Qiaquick PCR purification Kit (Qiagen) and labeled using the DIG-High Prime DNA Labeling Kit (Sigma-Aldrich). Pre-hybridization and hybridization were performed with DIG Easy Hyb solution (Sigma-Aldrich) at 42°C. Washes were performed for 10 min at room temperature in 2X SSC, 0.1% sodium dodecyl sulfate (SDS), and for 30 min at 68°C in 0.1X SSC, 0.1% SDS. NBT-BCIP (Sigma-Aldrich) was used for the detection.

The plasmid sequences adjacent to the 360-bp tetK fragment have been recovered by inverse PCR using inversely oriented primers, IP-AM1 AAAAGATAATCCGCCCATAACA and IP-AV1 TGCTTCTGGAATGAGTTTGCT. Approximately 1 ng of plasmid content was used as a template. Inverse PCR was performed in a 50 μl reaction mixture containing 5 pmol of each primer, 200 μM of each dNTP, 1× AccuTaq buffer, and 2.5 U of AccuTaq LA DNA polymerase (Sigma-Aldrich). The amplification cycle was as follows: 5 min at 94°C; 30 cycles of 30 s at 94°C, 30 s at 55°C, and 5 min at 68°C; and a final extension of 10 min at 68°C for amplification of an expected product of about 5 kb.

The amplification product was analyzed on 0.8% agarose gel and visualized by ethidium bromide. Finally, the inverse PCR product was gel-purified with extraction from agarose using the QIAquick Gel Extraction Kit (Qiagen) for primer-walking-strategy DNA sequencing. Sequences were assembled into a contig using CAP3 (Huang and Madan, 1999) and then analyzed using the NCBI BLAST suite of programs.

Genomic DNA of S. xylosus was prepared in agarose plugs as described previously (Morot-Bizot et al., 2003). The enzyme SmaI was used according to the manufacturer’s instructions (Promega). Digested DNA was subjected to pulsed-field gel electrophoresis (PFGE) in 1% agarose gels in 0.5× TBE buffer on a CHEF-DR III apparatus (Bio-Rad). Electrophoretic conditions were 50–100 s for 6 h and 10–30 s for 18 h at 14 C at 6 V/cm and an angle of 120°. Lambda DNA concatemers were used as molecular size markers (Promega). Gel was stained in ethidium bromide.

The GenBank accession number of the full-length sequence of plasmid pSX10B1 with annotations is MK433518.

The three S. xylosus strains resistant to tetracycline were used as donors, and the C2a strain was used as recipient in filter mating experiments. The transfer of tetracycline resistance was only observed in the case of a donor/recipient ratio of 1:10 and the XIV 10B1 strain as a donor. The transfer occurred at a low frequency of 3.4 × 10⁻⁹ per recipient, and two presumptive transconjugants (TC1 and TC2) were isolated.

To study the potential transfer of tetracycline resistance of the XIV 10B1 strain in situ in meat matrix, fermented sausages were made on a pilot scale mimicking the industrial process. Two batches were prepared with the same raw materials and inoculated with L. sakei as starter (sensitive to tetracycline), but with two different levels of S. xylosus (donor and recipient, ratio 1:1): one at approximately 6 log CFU/g (batch E1) and the other at 8 log CFU/g (batch E2) (Table 1). The ratio 1:1 has been chosen because plasmid transfers were observed for Lactobacillus curvatus during sausage fermentation (Vogel et al., 1992). The fermented sausages were analyzed over a ripening period of 30 days.

A growth of approximately two logs was recorded for lactic acid bacteria in both batches (Table 1). They grew during the fermentation and ripening steps. A drop in pH was noted during the fermentation phase followed by an increase during the ripening step in the two batches. Finally, water activity decreased during ripening in both batches (Table 1).

Enumeration of staphylococci on Chapman medium is shown in Table 1. The staphylococcal count remained close to their level of inoculation throughout the process in both batches. All randomly selected colonies were identified as S. xylosus. No transconjugants were isolated by replica plating of colonies grown on Chapman agar.

In batch E1, the donor strain grew during the fermentation step and then remained at the same level, while the recipient strain decreased slightly during the process (Table 1). In the condition of batch 1, no presumptive transconjugant was isolated on medium supplemented with Tet, Rif, and Fus (Table 1).

In batch E2, the donor and recipient strains remained at their level of inoculation up to the fermentation, and a slight decrease was noted at 30 days of ripening. Presumptive transconjugants were isolated on medium supplemented with Tet, Rif, and Fus after 6 days of fermentation (E2–1, E2–2, and E2–3) and after 30 days of ripening (E2–4). These presumptive transconjugants were only detected on plates inoculated from the 10⁻¹ sample dilution. The frequency of the transfer can be estimated at a rate of 1.4 × 10⁻⁷ per recipient after 6 days of fermentation.

The six colonies of presumptive transconjugants were picked from selective plates and were characterized to determine if these isolates were “true” transconjugants and not mutants of donor or recipient.

Antibiotic susceptibility for presumptive transconjugants and donor and recipient strains was determined by a disk diffusion test (Table 2). All transconjugants were resistant to tetracycline and sensitive to minocycline, like the donor strain. Their resistance to rifampicin and fusidic acid was identical to those of the chromosomally resistant recipient strain. The resistance to penicillin G of the donor strain had not been transferred. Note that the transferred tetracycline resistance is stably maintained in all transconjugants in the absence of selective pressure.

We have shown that PCR amplification was successful in detecting tetK in all presumptive transconjugants as in the XIV 10B1 donor strain (Figure 1). The sequencing of the XIV 10B1 360 bp-PCR product confirmed that the PCR was specific.

When comparing plasmid carriage of presumptive transconjugants with the donor strain, we have shown that all transconjugants harbored a small plasmid (<5 kb) of similar size to that of the donor strain (Figure 2A). This small transmissible plasmid hybridized with the tetK probe (Figure 2B). We have observed in the four transconjugants obtained in situ the additional transfer of a smaller plasmid (Figure 2B). The complete nucleotide sequence of the plasmid carrying tetK was obtained. Plasmid pSX10B1 is 4,498 bp long with 30.1% G + C content. The sequence encodes three open reading frames, which correspond to a replication initiation protein, a tetracycline resistance protein (TetK), and a plasmid recombination enzyme. It exhibits a high identity (96.5%) with the S. aureus COL plasmid pT181. Moreover, the amino acid sequence of TetK shows 100% identity with that of pT181.

Transconjugants, donor, and recipient were subject to PFGE analysis. The donor and recipient produced distinct patterns (Figure 3). The transconjugants displayed the same pattern as the recipient strain.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.