L. rhamnosus (ATCC 53103) was purchased from the American Type Culture Collection (Manassas, VA, USA). The bacteria were maintained on de Man, Rogosa and Sharpe (MRS) agar plates (Difco; BD Biosciences, Franklin Lakes, NJ, USA) and cultured overnight in MRS broth (Difco; BD Biosciences) at 20°C with gentle agitation. S. aureus (ATCC 29213; American Type Culture Collection) was used in the present study to screen for the antibacterial activity of bacteriocins from L. rhamnosus. Cultures of S. aureus were grown overnight in brain-heart infusion broth (Difco; BD Biosciences) at 37°C, centrifuged (3,000 × g at 4°C for 5 min), and the final concentrations of the cultures were adjusted to match the turbidity of a McFarland 0.5 standard (0.5×10⁵ CFU/ml) using a spectrophotometer (Densimat; BioMérieux UK Ltd., Basingstoke, Hampshire, UK).

Crude bacteriocin preparations were prepared as previously described (8). Briefly, L. rhamnosus cultures were initially grown in MRS broth at 37°C for 24 h. The cultures were then centrifuged (7,000 × g at 4°C for 10 min) and cell-free supernatants were then aspirated. The pH of the supernatants was adjusted to 6.5, treated with catalase (5 mg/ml) and filtered through 0.22 µm pore size filters (EMD Millipore, Billerica, MA, USA).

Crude bacteriocin preparations were saturated with 70% ammonium sulphate at 4°C to precipitate the proteins. The pellets were recovered by centrifugation at 10,000 × g at 4°C for 30 min, vacuum-dried and dissolved in 1 ml deionized water. The 10 ml suspensions were then applied to Sephadex G-100 columns (1.6×36 cm) (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) that were pre-equilibrated with a phosphate buffer (pH 7.0), as previously described (9). The flow rate was adjusted to 24 ml/h, and the proteins were collected and adjusted to a total protein concentration of 2.5 mg/ml. The protein suspension was stored at −70°C until required. A bacteriocins suspension (10 µl; 1 mg/ml) was loaded into a 12.5% Tricine-SDS-PAGE and the molecular weight of the bacteriocins was determined according to the methods of Biosa G (10). A total of 10 µl of the low molecular weight (10–120 kDa) BenchMark marker (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) was used as a standard.

A total of 12 New Zealand White female rabbits provided by the Zhejiang Province Academy of Medical Sciences (Hangzhou, China) weighing 2.5–3.0 kg and aged 70–100 days old were used in the present study. Animals were housed in individual cages in a temperature-controlled room (23°C) with a 12 h light/dark cycle. A total of 6 rabbits were used for validation of the infection model, while the other 6 were assessed as part of the control group. All animals were handled in strict accordance with good animal practice as defined by the National Institutes of Health Guide for the Care and Use of Laboratory Animals (11) and all animal work was approved by the College of Medicine of Ningbo University Chancellor's Animal Research Committee (Yuyao, China).

The right knee of each rabbit was replaced with a tibial component using a Silastic implant (Dow Corning, Midland, MI, USA). The surgery was performed as described previously (12). Briefly, the rabbits were anesthetized via inhalation of 2% isoflurane, which was maintained by the intramuscular injection of ketamine (25 mg/kg) followed by continuous inhalation of 1% isoflurane. The right leg of each rabbit was shaved 24 h prior to the surgery and the skin was cleaned with an iodine solution just before surgery. A longitudinal skin incision was made to expose the knee. The epiphyseal plates were removed following dislocation of the tibia. The metaphysis was exposed and the cancellous bone of the medullary cavity of the proximal metaphysis was reamed. The stem of the nail-shaped silicone implant (14 mm long) was inserted into the intramedullary canal of the tibia, such that the implant head (15×5 mm) replaced the tibial plateau. The deep fascia and skin were finally closed to complete the procedure.

Immediately after the surgical wound was closed, the 12 rabbits were infected by the intra-articular injection of 0 5 ml inoculum containing 3×10⁵ CFU/ml S. aureus. Then, 6 rabbits were injected with 1 ml of the bacteriocin suspension 4 h after the injection of S. aureus, and the other 6 rabbits were injected with 1 ml sterile saline solution.

To test for biofilm formation in the rabbit model used in the present study, the implants were harvested from euthanized rabbits 5 days after surgery. The specimens were fixed with 2.5% glutaraldehyde at 4°C for 1 h. Following fixation, the tissues were stained with 0.01% acridine orange solution at 4°C for 2 h (Sigma-Aldrich; Merck KGaA), as described previously (13), and washed three times with PBS. The 50-µm-thick sections were then observed under an Eclipse 80i microscope (Nikon Corporation, Tokyo, Japan) equipped with an argon laser at an excitation wavelength of 488 nm. All images were quantified using Image-Pro Plus software (version 6.0; Media Cybernetics, Inc., Rockville, MD, USA). The fluorescence intensities of the biofilms were expressed as integrated optical density (IOD) values, which were calculated by the multiplications of the area and the density of the biofilm (14).

Blood samples were collected at 1, 2, 3, 4 and 5 days after the injection of bacteriocin or saline. The samples were centrifuged at 7,000 × g at 4°C for 5 min and then stored at −70°C until required. CRP and IL-6 levels were measured using commercial ELISA kits for CRP (cat. no. E-EL-RB0005) and IL-6 (cat. no. E-EL-RB0014; both Elabscience Biotechnology Co., Ltd., Wuhan, China), respectively, according to the manufacturer's protocol All measurements were performed in triplicate. The levels of CRP and IL-6 were expressed as mg/ml protein.

All tests were repeated three times. The results are presented as the mean ± standard deviation. Data analysis was performed using SPSS software (version 14.0; SPSS, Inc., Chicago, IL, USA). The results between two groups were compared using a Student's t-test. P<0.05 was considered to indicate a statistically significant difference.

The IOD values of the biofilms the 6 infected rabbits revealed that there was a significant decrease in biofilm formation compared with the control group 5 days after surgery and infection (P<0.01; Fig. 1). Fluorescence microscopy of the biofilms revealed that the density of bacteria in the rabbits injected with the bacteriocin suspension was markedly lower than that of the control group (Fig. 2). The Tricine-SDS-PAGE of the bacteriocin suspension is presented in Fig. 3. The molecular weight of the bacteriocin was found to be 10–15 kDa.

Due to the infection with S. aureus following the surgery, the levels of CRP and IL-6 in the control group increased, reaching maximum levels at 5 days post-surgery (Fig. 4). However, the levels of CRP and IL-6 in the rabbits treated with bacteriocin remained at a steady low level, and were significantly lower than the levels in the control group at days 3, 4 and 4 post-surgery (all P<0.05; Fig. 4).