Bacterial culture
MDR S. aureus isolate 63-2498 (MDR SA63-2498) collected from an Adelaide DFU patient, was identified using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI TOF MS) (Bruker Daltonics Biotyper, Bruker Pty. Ltd., Victoria, Australia). MDR SA63-2498 was tested using a VITEK® 2 system (bioMérieux, New South Wales, Australia) and demonstrated resistance to multiple antibiotics, including benzylpenicillin, oxacillin, ciprofloxacin, erythromycin, and clindamycin and susceptibility to vancomycin. MDR SA63-2498 was susceptible to lysis by AB-SA01 and each of its component phages in planktonic and biofilm states during the pilot study. A single colony of MDR SA63-2498 grown on a mannitol salt agar (MSA) (Thermo Fisher Scientific, South Australia, Australia) plate was taken and grown in 3 ml trypticase soy broth (TSB) (Thermo Fisher Scientific, South Australia, Australia) overnight. One milliliter from the TSB overnight broth culture was adjusted to an optical density (OD) of 0.7 at OD600 using a SP-830+ Metertech spectrophotometer (Adelab, South Australia, Australia) that corresponds to 2.3 log10 (CFU/ml) through dilution with sterile phosphate-buffered saline (PBS). The adjusted broth culture was centrifuged at 3000 rpm, and pelleted cells were washed twice and resuspended with 1 ml PBS. This suspension was used for inoculation into mice’s wounds.
Phage cocktail
Phage cocktail AB-SA01 was provided by AmpliPhi Biosciences Corporation. It is a combination of three Myoviridae phages designated J-Sa36, Sa83, and Sa87. The mean titer was 9.3 log10 (PFU/ml) for J-Sa36 and Sa83, 9.0 log10 (PFU/ml) for Sa87, and 9.1 log10 (PFU/ml) for AB-SA01 on S. aureus laboratory strains RN4220 and SA6538 using plaque assay (22).
Laboratory animal management
Female Balb/c mice were obtained from the Animal Resources Centre, Perth, Western Australia. All experiments were approved by the Animal Welfare Committee, Flinders University/Southern Adelaide Local Health Network, and carried out in compliance with the ARRIVE guidelines (43). Mice were kept at the College of Medicine and Public Health Animal Facility, Flinders University, at 22 ± 3 oC and 55 ± 5% humidity under a 12:12 hour light-dark cycle in Tecniplast GM500 Mouse IVC Greenline cages (Tecniplast Australia Pty Ltd, New South Wales, Australia) using corncob bedding. Mice were kept in specific pathogen-free conditions. Mice were provided water and meat-free rodent maintenance diet (Glen Forrest Stockfeeders, Western Australia, Australia) ad libitum. The research was conducted consistent with the Australian Code for the Care and Use of Animals for Scientific Purposes, 8th edition, 2013.
Mice were weighed 2-3 times each week and monitored for hunching, ruffled coat, lethargy, cold to touch, crinkling of skin, sunken eyes, and rapid or labored breathing at least once daily. Mice that showed over 15% weight loss or were critically ill were euthanized. Vancomycin (Sigma-Aldrich Corporation, New South Wales, Australia) was assessed for possible toxicity on six mice in a separate pilot study at 150 mg/kg dose rate, twice daily IP, for five consecutive days as described (44); these mice did not display any apparent adverse reaction. At the end of the experiment, mice were euthanized using 3% isoflurane, followed by cervical dislocation. Post-mortem examination of the external surfaces and visceral organs was conducted following an established procedure (45).
Induction of diabetes in mice
The diabetes mouse model was used to mimic the human diabetes setting, where the experimental bacteria are collected. A total of 48 female 8-week-old Balb/c mice, housed in groups of 5, received streptozotocin (STZ) (Sigma-Aldrich Corporation, New South Wales, Australia) following an established protocol (24). Balb/c mice are among the least susceptible to STZ toxicity because of their relatively high pancreatic β-cell mass due to their large number of islets (46). Female mice are relatively resistant to the glucotoxicity of STZ compared to males because sex steroids protect them from β-cells injury (47). Besides, STZ-diabetic nephropathy is more pronounced in male mice compared to females (46). STZ is a naturally occurring alkylating antineoplastic agent that is toxic to pancreatic insulin-producing cells. It is used to treat pancreas islet cell carcinoma and to induce diabetes mellitus in laboratory rodents (23).
After 4 hours of fasting, each mouse received an IP injection of 50 mg/kg freshly dissolved STZ in 0.05M citrate buffer pH 4.5 once daily for five consecutive days. Non-fasting blood glucose levels (BGL) were checked every 2 - 3 days by tail vein bleeding after applying 3% lidocaine/prilocaine anesthetic cream. BGLs were measured using an Accu-Chek Performa (Roche Diabetes Care Australia, New South Wales, Australia) blood glucose meter. Mice with non-fasting BGL < 13.9, between 13.9 and 22.2, and ≥ 22.2 mmol/L on at least two different days were categorized as normal glycaemic, moderate hyperglycaemic, or severe hyperglycaemic status, respectively, as described earlier (48). Mice were treated with 1.0 IU NovoMix® 30 (Novo Nordisk, New South Wales, Australia) insulin subcutaneously daily, from the second week of diabetes manifestation, to ameliorate the hyperglycaemic effects of diabetes and maintain body weight (49).
Excisional wound infliction
Diabetic mice were provided with lemon-flavored paracetamol in drinking water at 1.34 mg/ml 3 days prior to wound infliction and for the entire experimental period. An established rodent wound infection model (27) was used. Mice were anaesthetized using 3% isoflurane inhalation and maintained on 1.5% during surgery. Hair was removed on the dorsal skin using electric clippers and depilatory cream, and skin sterilized using 70% ethanol. Mice were given a single 0.05 mg/kg buprenorphine injection subcutaneously before the incisions. Hydrating eye drops were applied during anesthesia.
A skin wound extending through the Panniculus carnosus muscle was inflicted using a 6 mm sterile biopsy punch and fine scissors. The bilateral wounds were at 10 mm either side of the midline and 30 mm from the base of the skull. Swab samples were collected from each wound to assess for existing S. aureus. A sterile 1 mm thick silicone splint with a 7 mm diameter circular hole at the center was applied over the wound and sutured onto the skin concentric with the wound using 5-0 nylon suture. Cyanoacrylate glue was used to fix the silicone splint to the skin before suturing. The silicon splints were removed at the end of the experiment.
Wound infection and treatment
Mice were randomly assigned into S. aureus-inoculated phage-treated (n = 8), S. aureus-inoculated vancomycin-treated (n = 6), S. aureus-inoculated PBS-treated (n = 7), and PBS-inoculated phage-treated (n = 8) groups. S. aureus-inoculated mice were infected with 50 µl suspension containing 6.7 log10 (CFU) of MDR SA63-2498 prepared as above. Mice in the PBS-inoculated group received 50 µl of PBS. Suspensions were applied directly into each wound, overlaid with gauze, and covered with Opsite sterile transparent wound dressing, which adhered to the silicon splint and retained in place the gauze dressing. Post-infection, each mouse was kept in a single housing system. On days 3, 5, and 7 post-wounding, the Opsite was removed, and swab samples were taken. Swab samples were taken by scrubbing the surface of each wound by rotating three times clockwise with enough pressure to produce a small amount of exudate and inserted into a separate tube of 1 ml TSB. The tube was vortexed (with the swab inside) for 5 seconds, and a 100 µl aliquot was used for 10-fold serial dilutions. Swab samples were kept at +4 oC and processed for the bacterial count within 4 hours of collection.
Treatments were administered after sample collection commencing on day 3. Gauze patches (10 x 10 mm) soaked with 70 µl AB-SA01, equivalent to 7.9 log10 PFU, or 70 µl PBS solutions for control and vancomycin-treated mice, were applied to wounds and covered with Opsite, on days 3, 5, and 7. Vancomycin-treated mice received 150 mg/kg vancomycin IP twice daily for five consecutive days, as described (44). Topical vancomycin was not an option because its best topical dose is unknown, it has spoor tissue penetration, and could contribute to vancomycin resistance (28, 50).
Assessment of treatment effect on bacterial load and wound healing
Wound size was measured in duplicate from multiple directions using a digital Vernier caliper by tracing the leading edge of epithelium within the wound (51). The wound size on the excision day was defined as the original wound size. One hundred microlitres of 10-fold serially diluted swab sample suspensions were mixed with 3 ml trypticase soy soft agar and cultured on MSA to assess the bacterial load. After 24 hours of incubation at 37 oC aerobically, colony counts were performed on plates with 30-300 colonies as recommended (52). The bacterial population was calculated using the formula B = N/d where B = number of bacteria, N = average number of colonies, and d = dilution factor.
Data management and statistical analysis
Data were double entered, encoded, and stored using Microsoft Excel Spreadsheet. STATA (version 16) software was used for statistical analysis. Data are reported in terms of mean ± standard error. CFU data are expressed as logarithm-transformed values (log10 (CFU /ml)) over time. A comparison of experimental groups was performed using a one-way analysis of variance (two-tailed) or paired ‘t-test’. A p < 0.05 value was considered statistically significant.