Bacterial strains, culture and antibiotics
The clinical S. aureus isolate SAU060112 was used for both in vitro and in vivo experiments. SAU060112 was originally isolated from a prosthetic knee infection21. S. aureus cultures were cultivated on tryptic soy agar (TSA) and grown in tryptic soy broth media (TSB) (Sigma-Aldrich) overnight for 16–20 h (h) at 37˚C, 180 rpm. Additionally, a coagulase-deficient S. aureus double-knockout mutant (S. aureus ATCC 29213 ∆vWbp ∆coa) was used for in vitro experiments that involved human plasma in TSB media22. Modified M9 buffer (mM9) (Sigma-Aldrich) was used for bacterial starvation23. mM9 contains M9 minimal salts (x1), 0.1 mM CaCl2, 2 mM MgSO4, 1 mM Thiamine-HCl, 0.05 mM nicotinamide, and 1 ml/L TMS324. M9 solutions were adjusted to pH = 7.4 before autoclaving. Vitamins, salts and trace metals were sterile filtered and added to autoclaved M9 solution23. For all in vitro experiments involving daptomycin, 50 mg/L CaCl2 was added to the specific medium.
Minimum inhibitory concentration (MIC) and minimal bactericidal concentrations (MBC)
MIC was determined by broth microdilution. Three overnight cultures of S. aureus were prepared and subsequently adjusted to OD600 = 0.01 (5 × 105 CFU/ml). Compounds were dissolved in DMSO (max 1%) or NaCl (0.9%) and diluted in two-fold series in a 96-well microtiter plate. After 24 h of treatment at 37˚C, 50 rpm wells were visually inspected, and growth was measured by optical density (OD600). Growth inhibition was determined by no visible growth in wells and < 20% OD-value relative to OD-value of growth control. MBC was determined as the concentration with < 5 colonies after spotting out 10 µL from wells onto TSA and 24 h incubation at 37˚C.
Checkerboard assay
Possible synergy or antagonism between compounds were tested using a checkerboard assay. Overnight cultures of S. aureus were adjusted to OD600 = 0.01 (5 × 105 CFU/ml). The checkerboard assay was performed in a 96 microtiter well containing TSB and two-fold-dilutions of bithionol horizontally and either moxifloxacin or daptomycin in two-fold serial dilutions laterally. Bacteria were then added to all wells and allowed to grow for 24 h at 37˚C, 50 rpm before visual inspection and OD600 measurement. Experiments were performed in triplicates. Synergism, no interaction or antagonism was determined by calculating the Fractional Inhibitory Concentration Index (FICI). FICI 0.5-4.0 indicated additive/indifferent interaction, FICI < 0.5 indicated synergy whereas FICI > 4.0 indicated antagonism25.
Minimal biofilm eradication concentration (MBEC)
Four S. aureus overnight cultures were prepared and adjusted to OD600 = 1 (5 × 107 CFU/ml). PEG-lids (NUNC™ Immuno TSP lids, ThermoFischer Scientific (cat#445497)) were inoculated in adjusted culture for 60 min, 37˚C, 50 rpm. PEG-lids were then transferred to fresh TSB every 24 h and incubated at 37˚C, 50 rpm for a total of 48 h maturation before treatment start. Biofilm bacteria were treated 24 h with antibiotics dissolved in either DMSO or NaCl (0.9%) in a two-fold dilution series (concentration ranges of bithionol, daptomycin and moxifloxacin were 128 − 0.25 mg/L, 256 − 0.5 mg/L and 32-0.063 mg/L respectively and as for combination therapy experiments, bithionol concentration ranged from 32 to 0.063 mg/L whereas antibiotics daptomycin and moxifloxacin were held constant at 4 mg/L and 1 mg/L, respectively) at 37˚C, 50 rpm. PEG-lids were washed 2 × 1 min in fresh TSB media before sonication at 45 kHz for 10 min to release the biofilm. Biofilm bacteria were allowed to recover and grow for 24 h at 37˚C, 50 rpm before identifying the minimal biofilm eradication concentration (MBEC) as the lowest antibiotic concentration that did not lead to recovery of bacterial growth.
Antimicrobial activity against persister cells
The persister cell assay has previously been described23. Briefly, three overnight cultures were prepared and subsequently diluted 1:1000 in fresh TSB and grown 16–20 h at 37˚C, 180 rpm. Cultures were washed by centrifugation at 13.150 x g for 10 min and resuspended in mM9 buffer, thus forcing the bacteria to starve to induce persister formation. Cultures were then adjusted to OD = 1 (5 × 107 CFU/ml) in mM9 buffer. 20µL of washed culture were added to a microtiter plate, containing mM9 buffer and either bithionol at 0.01 ×, 0.05 ×, 0.1 ×, 0.25 ×, 0.5 ×, 1 ×, 2 × and 4 × MIC or in combination with antibiotics daptomycin or moxifloxacin at concentrations 0 ×, 5 × and 10 × MIC. The bacteria were treated 24 h at 37˚C. CFU was done from the three washed cultures before addition of antibiotics (T0) and after 24 h (T24). To wash off antibiotic remnants, 50 µL samples were centrifuged (as previously described) and resuspended in 50 µL mM9 buffer before 10-fold dilution series and plating on TSA were performed. CFU enumeration was done after 24 h incubation at 37˚C.
Ethical statement
The animal studies were approved by the Animal Research Inspectorate under the Danish Ministry of Justice (Permission number: 2022-15-0201-01133) and followed the Danish act on animal experiments (LBK nr 474 of 15/05/2014 and BEK nr 2028 of 14/12/2020) as well as the EU directive 2010/63/EU on the protection of animals involved in scientific research. All animal experiments were performed in accordance with the relevant guidelines, including the ARRIVE 2.0 guidelines and the principles of the 3Rs. Animals were closely monitored with evaluation of activity levels and weighed every fourth day to prevent and minimize suffering (Supplementary Fig. 1). Animals were euthanized if humane endpoints were reached. Studies were conducted in close collaboration with veterinarians at the animal facilities at Skou building, Aarhus University. The study was not blinded.
Antibiotics used in vivo
Bithionol was dissolved in Kolliphor HS-15 (Sigma-Aldrich)/ethanol 96%, 1:1 v/v and diluted 1:10 in saline to a concentration of 6 mg/ml = 30 mg/kg dose9. Daptomycin (“Cubicin”, Merck Sharp and Dohme) and Moxifloxacin (KRKA) were dissolved in NaCl (0.9%). Concentrations of daptomycin and moxifloxacin were 10 mg/ml and 1.6 mg/ml respectively which yielded doses of 50 mg/kg and 10 mg/kg respectively.
Inoculation of tibial steel implants
Stainless steel insect pins (Benfidan, Nykøbing Mors, Denmark) were used as implants. An overnight culture was prepared and subsequently adjusted to OD600 = 0.1 (5 × 106 CFU/ml) in TSB. Pins were placed in a 15 ml falcon tube, each containing 5 pins and 5 ml of adjusted S. aureus culture. Pins were followingly incubated stationary at 37oC for 24 h until use for surgery the next day.
Implant associated osteomyelitis in vivo model in mice
60 6–8-week-old female C57bl/6j mice (Janvier Labs, Le Genest-Saint-Isle, France) were included in the study. At the time of inoculation, weight was between 17 and 21 g. Mice were acclimatised for a minimum of 7 days and subsequently housed at the animal facility at Department of Biomedicine, Aarhus University.
The surgical procedure was based on a model by Jørgensen et al.5. Briefly, mice were given analgesics prior to surgery (buprenorphine 0.05 mg/kg s.c.) and then anesthetized with isoflurane (induction 5%, maintenance 2%), oxygen (0.5 L/min) and air flow (1.2 L/min). The left hind leg was shaved, disinfected and the implant was pushed through the tibia at the proximal epiphysis. After insertion, the implant was bent in a U-shape and cut as close to the skin as possible. The skin and adjacent tissue was then managed to cover the implant ends. Buprenorphine (0.7 mg/kg) was administered in drinking water until 4 days post-surgery.
Following 7 days of infection, mice were randomised into six treatment groups with 10 mice in each group (pr. cage) by letter randomization. Groups were: (1) NaCl (0.9%) control (once daily); (2) bithionol (30 mg/kg/12 h); (3) bithionol (30 mg/kg/12 h) + daptomycin (50 mg/kg/24 h); (4) bithionol (30 mg/kg/12 h) + moxifloxacin (10 mg/kg/12 h); (5) daptomycin (50 mg/kg/24 h); (6) moxifloxacin (10 mg/kg/12 h). Treatment duration was 5 days and all drugs were administered intraperitoneally. No animals died or were excluded during the experiment. 36 h after last treatment dose, the mice were sedated and euthanized by cervical dislocation. Tibias including implants were carefully removed and kept on wet ice until sample preparation for bacterial quantification was performed immediately after.
Post-mortem bacterial quantification
Implants were delicately separated from the dissected tibial bone and placed in Eppendorf tubes with 1 ml phosphate buffered saline (PBS). The tibial bone was cut in smaller fragments and transferred to Precellys MK28 hard tissue grinding tubes (Bertin Technologies, Saint Quentin, France) with 1 ml PBS and kept on wet ice. Bones were homogenized for 2 × 20 s at 5000 RPM in a bead beater (MagNA Lyser Instrument, Roche Diagnostics), transferred to wet ice to cool down for 2 min and homogenized again. Meanwhile, implants were sonicated in an ultrasonic bath (USCS1700 T, VWR, Westchester, PA, USA) for 5 min at 45 kHz and 110 W to release the biofilm on the implants. Implants were vortexed thoroughly in PBS before and after sonication. Now, 10-fold dilutions series of both sonicate and bone homogenate were made in PBS and 10 µl from each dilution row was plated out on TSA and incubated at 37oC for 24 h before enumerating CFU of all samples.
Statistical analysis
Normality of data was assessed by QQ-plots. Parametric data were compared with either Student’s t-test or one-way ANOVA with Bonferroni post hoc correction and presented as means ± SD. Statistical analysis and graphs were made with GraphPad Prism (10.2.3, GraphPad software, San Diego, California, USA).
