Treating methicillin-resistant Staphylococcus aureus (MRSA) bone infection with focused ultrasound combined thermally sensitive liposomes

Abstract Objective Chronic bone infection caused by Staphylococcus aureus biofilms in children and adults is characterized by reduced antibiotic sensitivity. In this study, we assessed ‘heat-targeted, on-demand’ antibiotic delivery for S. aureus killing by combining ciprofloxacin (CIP)-laden low-temperature sensitive liposomes (LTSLs) with local high-intensity focused ultrasound (HIFU) induced bone heating in a rat model of bone infection. Methods CIP-LTSLs were prepared using the thin-film hydration and extrusion method. Bone infection was established by surgically implanting an orthopedic K-wire colonized with methicillin-resistant S. aureus (MRSA) strain into rat’s femurs. For bone heating, ultrasound-guided HIFU exposures were performed to achieve a local temperature of 40–42 °C (∼15 min) concurrently with intravenous injection of CIP-LTSLs or CIP. CIP biodistribution was determined spectrophotometrically and therapeutic efficacy was determined by bacteriological, histological and scanning electron microscopy (SEM) analyses. Results CIP-LTSLs in the range of 183.5 nm ± 1.91 showed an encapsulation efficiency of >70% at 37 °C and a complete release at ∼42 °C. The metal implantation method yielded medullary osteomyelitis characterized by suppurative changes (bacterial and pus pockets) by day 10 in bones and adjoining muscle tissues. HIFU heating significantly improved CIP delivery from LTSLs in bones, resulting in a significant reduction in MRSA load compared to HIFU and CIP alone groups. These were also verified by histology and SEM, wherein a distinct reduction in S. aureus population in the infected metal wires and tissues from the combinatorial therapy was noted. Conclusion HIFU improved CIP delivery to bones, achieving clearance of hard-to-treat MRSA biofilms.


Introduction
Osteomyelitis is a clinical complication that occurs due to the formation of bacterial biofilms in bone, bone marrow and surrounding tissue.Biofilm-associated osteomyelitis is highly resistant to antibiotics, often requiring multiple revision surgeries and resulting in functional impairments and amputations [1,2].Antibiotic resistance, including that related to osteomyelitis, is projected to cause 10 million deaths/year by 2050.The treatment of implant-associated infection is costly, with estimates exceeding $1.62 billion by 2020 [3].The risk of reinfection increases after every revision surgery, as the bacteria within the biofilms exhibit altered growth rates, gene expression profiles and protein synthesis [3,4].These changes make the bacterial pathogens highly resistant to the effector mechanisms of the host's immune system.Therefore, there is an urgent need to develop innovative and novel treatment approaches that can address the complex variables of osteomyelitis biofilms for improved therapeutic outcomes.
In all osteomyelitis treatment scenarios, success depends on the complete removal of infected devitalized tissues and metallic hardware.An essential factor also governing infection management is the antibiotic targeting of the pathogens.In this regard, the prophylactic and therapeutic use of antibiotic-loaded implant depots and biodegradable spacers have been attempted; however, they have had only moderate clinical success due to issues with sub-therapeutic delivery and drug resistance [3,5,6].As an alternative, we propose a combination of high-intensity focused ultrasound (HIFU) therapy and low-temperature sensitive liposomes (LTSLs) to deliver antibiotics.Our LTSLs release encapsulated antibiotics slightly above body temperature (42 C) [7], and we hypothesized that systemically administered antibiotic-laden LTSLs can be activated by HIFU heating to achieve chemotherapy for implant-associated osteomyelitis.
The clinical efficacy of HIFU in achieving deep-seated heating noninvasively has been well established [8].HIFU exposure parameters can be manipulated to attain thermal and/or mechanical effects, which can be particularly effective in targeting biofilms.Additionally, HIFU can enhance the delivery and passage of drugs and immune cells, as well as increase cell sensitivity to chemotherapy [9].Our previous work has demonstrated the effectiveness of HIFU and ciprofloxacin (CIP)-loaded LTSLs in managing superficial chronic wounds [7].However, it is unknown whether this approach can effectively target hard-to-reach infected marrow regions of bones and improve antibiotic therapy.To address this, in this study, we tested the combined approach of CIP-LTSL and HIFU treatment in a rat model of chronic implant-associated osteomyelitis, assessing localized antibiotic delivery and efficacy (Figure 1).Our findings suggest that this minimally invasive and non-surgical HIFU-based approach for treating recurrent biofilm infections can potentially improve outcomes in patients.

Synthesis of LTSLs and CIP loading
For LTSLs preparation, a thin film of phospholipids was hydrated and then extruded through a polycarbonate membrane filter [10] (Figure 2).Briefly, DPPC, LysoPC and DSPE-mPEG were dissolved in chloroform at a molar ratio of 85.3:9.7:5.0.The organic solvent was evaporated using a rotary evaporator, and the resulting thin lipid film was hydrated using 350 mM ammonium sulfate.Hydrated lipids were extruded five times through double stacked 200 nm polycarbonate filters to yield blank LTSLs.CIP loading (5 mg of CIP per 100 mg of lipids) into LTSL (CIP-LTSL) was carried out using the ammonium sulfate gradient as described previously [11,12].Unencapsulated CIP was removed using a PD-10 desalting column equilibrated with 150 mM NaCl.

Characterization of CIP release from CIP-LTSLs
CIP-LTSLs were characterized for size and polydispersity index using dynamic light scattering (DLS) with a NanoBrook 90Plus PALS device (Brookhaven Instruments Corporation, Holtsville, NY).Briefly, 10 ml of CIP-LTSLs were added to 1 ml of distilled water in a cuvette, and DLS measurements were recorded at room temperature.On average, five measurements were taken, and mean size and standard deviation were calculated.To confirm the structural stability, LTSL size was measured daily for 14 days, stored at 4 C.For zeta potential measurements, 10 ml of CIP-LTSLs were added to 1.5 ml of distilled water in a cuvette, and five measurements on average were taken to calculate the mean zeta potential.CIP-LTSL morphology was confirmed with transmission electron microscopy (TEM).CIP release was measured spectrophotometrically for three days using a Cary Eclipse Fluorescence Spectrometer (Agilent Technologies, Santa Clara, CA).Briefly, 10 ml of CIP-LTSLs were diluted 300-fold in distilled water and 10X triton, and 3 ml of sample were placed in a quartz cuvette with a magnetic stirrer.The fluorescence intensity of the fully released CIP was measured at an excitation wavelength of 330 nm and an emission wavelength of 445 nm, over a temperature range of 25-45 C, to analyze the temperature-dependent drug release from LTSLs.The encapsulation efficiency of the CIP-LTSLs was assessed using our previously published protocol [7].

In vitro assessment of antibacterial activity of CIP-LTSLs
MRSA colonies grown in trypticase soy broth (TSB) overnight were added to fresh TSB.Once the culture optical density reached 0.4-0.6, the grown cultures were divided into the following groups and treated accordingly: (1) untreated control, (2) free CIP (2 mg/ml), (3) blank LTSL, (4) CIP-LTSL (2 mg/ml CIP), (5) HIFU alone, (6) CIP þ HIFU, (7) LTSL þ HIFU and (8) CIP-LTSL þ HIFU.Cultures were incubated for 18 h in an incubator shaker at 37 C, 200 rpm and CFU/ml was calculated for each treatment group post-spectrophotometric analysis.To verify therapeutic effects, we assessed bioluminescence in planktonic bacteria cultures treated as described above (Bruker In-Vivo Xtreme II, Billerica, MA).To determine the ability of LTSLs to penetrate MRSA biofilms grown in 12well plates for 72 h at 37 C in TBS media, DiO (DiOC 18 (3), green-fluorescent lipophilic carbocyanine dye) coated LTSLs were added to the biofilm at a concentration of 100 ml/ml of media for 24 h.Plates were imaged at 4 h, 8 h, 12 h and 24 h after carefully removing TBS and gentle PBS washing using Agilent BioTek Cytation 5 Cell Imaging Multimode reader (Santa Clara, CA).Images were captured under green channel and brightfield and presented as an overlay of both channels.

Establishment of implant MRSA biofilm-associated osteomyelitis in the rat model
All animal-related procedures were approved and carried out under the regulations and guidelines of the Oklahoma State University Animal Care and Use Committee.Briefly, 6-8week-old male Wistar rats (Charles River, Worcester, MA) were anesthetized with 1 l/min oxygen and 5% isoflurane induction, and a 3% maintenance dose.Meloxicam was administered subcutaneously at 1 mg/kg body weight.For metal implantation, the rat was placed in left lateral recumbency, and the surgical area around the femur was shaved and disinfected with chlorhexidine and iodine to create a sterile field.A medial incision was made along the knee joint using a scalpel blade (#11).The fascia was bluntly separated to provide better visualization of the medial patellar ligament, followed by medial parapatellar arthrotomy and lateral patella subluxation to expose the knee joint.An 18-gauge hypodermic needle was slowly inserted into the medullary cavity of both left and right femurs through the distal end of the femur, and replaced with a sterile orthopedic wire (22 gauge, approximately 2 cm long, made of stainless steel; IMEX Veterinary, Inc., Longview, TX) to fit inside the medullary cavity, while avoiding any irritation or infection in the knee joint.Next, 100 ml of MRSA, grown overnight in TSB, were centrifuged, washed, and diluted with 1Â phosphatebuffered saline (PBS) to a concentration of 1.5 Â 10 8 colonyforming units (CFU)/ml, and injected into the medullary cavity of the right femur through the distal end using an 18-gauge hypodermic needle (Figure 1(A)).After metal insertion and infection, the site was rinsed with sterile saline solution, and the incision was closed using simple interrupted intradermal skin sutures (4-0, PDS Ã II, Ethicon, Raritan, NJ) and tissue adhesive (3M Vetbond).The surgical region was sprayed with a taste deterrent spray (Grannick's bitter apple) to discourage rats from biting or licking the suture site.Finally, rats were X-rayed (Bruker In-Vivo Xtreme II, Billerica, MA) to confirm the successful implantation of the wire.All animals were given subcutaneous meloxicam injections for two days post-infection for pain management.

HIFU treatment setup and methodology
An integrated ultrasound-HIFU Alpinion system (VIFU2000, Bothell, WA) with a 1.5 MHz transducer frequency, 45 mm radius and 64 mm aperture diameter with a central opening of 40 mm in diameter was used for treatment planning and HIFU exposure.For in vitro treatment, MRSA cultures grown in thin-walled plastic tubes were mounted on the 3D positioning stage and treated for 120 s each, for three selected focal points, to cover more than 60% of the MRSA culture volume.We used the following HIFU parameters: 50% duty cycle, 5 Hz pulse repetitive frequency and 8 W power (equivalent to 3.6 W acoustic power).
For in vivo study, treatment was initiated 10 days after the onset of bone infection.Briefly, rats were anesthetized with 2-5% isoflurane, and a tail vein I/V catheter was secured prior to the rats being restrained in custom-built animal holders mounted on a 3D positioning stage.Then, they were lowered into a 37 C degassed distilled water bath for coupling with the rat hindlimb (Figure 3(A)).The right femur was aligned to the HIFU beam axis using real-time ultrasound guidance.VIFU software was used to define the target boundaries in the x, y and z directions for automatic rastering of the transducer.The following HIFU treatment parameters used were: 50% duty cycle, 5 Hz pulse repetitive frequency and 8 W power (equivalent to 3.6 W acoustic power).These parameters achieved a mean target temperature of 42-45 C at the bone-muscle interface, as measured using a temperature sensor inserted at the interface between the bone and muscle of the rat.Each focal point (1 Â 1 Â 10 mm on the x, y and z axes, respectively) within the raster treatment pattern was heated for 120 s, and the entire length of the femur was treated for a total treatment time of 30-35 min (Figure 3(B)).Although there may have been some exposure of bones at the interface with tissues, the animal skin remained intact post HIFU exposures.Additionally, distilled water was used for all HIFU experiments.They were not sterilized prior to the experiments since the infection was resident in the bones.CIP and CIP-LTSLs were injected intravenously at 10 mg/kg CIP concentration post-initiation of HIFU treatment.

In vivo CIP delivery and efficacy study design
To optimize infection rates, the rats were initially evaluated histopathologically at days 7, 10, 14 and 21 post-infection to track the progression of osteomyelitis.Histological and radiological features indicative of bone infection were observed at 10 days post-MRSA infection (Figure 4), and thus, this time point was selected for biodistribution and efficacy evaluations in the rats.Animals with osteomyelitis were randomized into the following groups: (1) control, (2) HIFU, (3) free CIP, (4) free CIP þ HIFU, (5) CIP-LTSL and (6) CIP-LTSL þ HIFU.For the drug delivery study, rats were sacrificed 24 h post-LTSL treatment, and CIP concentration was determined in the heated bone and adjoining muscle and compared between the free CIP ± HIFU and CIP-LTSL ± HIFU groups (n ¼ 3) (Figure 5(A)).To assess treatment efficacy, rats were given two treatments three days apart and sacrificed 24 h after the second treatment (Figure 5(C)).

Gait/pain analysis post-HIFU treatment
To assess the impact of HIFU treatment of bone on the ambulation of rats, the rats were scored on a scale of 0-4, where 0 indicated normal gait and behavior; 1 indicated the animal moving around while infrequently lifting the treated leg; 2 indicated the animal moving with a slight limp (partial weight-bearing); 3 indicated the animal was reluctant to move much with occasional setting down of the treated foot (toe touch); and 4 indicated a non-weight-bearing animal.The rats were monitored immediately post-treatment and at 6, 24 and 48 h post-treatment.

Post-treatment tissue analysis
Rats were euthanized 24 h post-treatment.The treated and untreated bone and surrounding muscles and skin were collected for bacteriological, CIP delivery and histological analysis.

Bacteriological assessment
For bacteriological analysis, the right femur bone and surrounding muscles and skin were collected, weighed and homogenized in 1Â PBS using a bead-vial homogenizer (Mini-Beadbeater-16, BioSpec, Bartlesville, OK) at 3450 oscillations/min for 3 min in 7 ml polypropylene screw-cap micro vials (BioSpec, Bartlesville, OK) using zirconia beads (1 mm diameter, BioSpec, Bartlesville, OK).Homogenized tissue samples were covered in foil and stored at À80 C before and after use.For analysis, 1:10 serial dilutions were made in 1Â PBS up to seven dilutions and plated on trypticase soy agar plates using the drop plate method.The resulting discrete colonies were counted to identify the highest dilution containing 30-300 colonies.The results were expressed as CFU/ml of homogenized tissue.

Analysis of CIP in tissues
To measure the amount of CIP in tissues, 100 mg of the homogenized tissues were added to 1.2 ml extraction medium (2% aqueous acetic acid/acetonitrile, 1:1 v/v) and mixed using the bead-vial homogenizer with zirconia beads.The sample lysate was transferred to 1.5 ml tubes and centrifuged to pellet cell debris at 16,000 Â g for 15 min.The supernatant was transferred to 1.5 ml tube and re-centrifuged.For CIP detection, 500 ml of clarified supernatant were added to a 700 ml quartz cuvette, and fluorescence (excitation wavelength: 330 nm, emission wavelength: 445 nm) was measured using a SpectraMax M2 spectrophotometer (Molecular Devices, San Jose, CA).

Histopathological assessment
For histopathological analysis, the surgical limb was removed, cut into half with the surrounding muscle, and stored in formalin.The paraffin embedded tissue sections were stained with hematoxylin and eosin and with Gram stain.The samples were scored for relative abundance of intra-osseous Gram-positive cocci and extent of tissue damage by an experienced pathologist blinded for this study.

Scanning electron microscopy (SEM) of S. aureus biofilms on explanted wires
For SEM analysis, the implants were carefully removed from the femur with sterile forceps and immediately fixed for 2 h in 2.0% glutaraldehyde in 0.1 M cacodylate buffer washed with sodium cacodylate buffer.Then, they were incubated for 1 h in 1% osmium tetraoxide in cacodylate buffer, serially dehydrated in increasing concentrations of ethanol (50, 70, 90, 95 and 100%), and dried in hexamethyldisilazane. Individual samples were mounted on microscopy stubs with tape, coated with gold-palladium, and viewed under the SEM to assess the biofilm architecture on the metal wire implants.

Statistical analysis
All analyses were performed using GraphPad Prism 9.0 (GraphPad Software Inc., San Diego, CA).All data were reported as the mean ± standard error of mean.A one-tailed t-test was performed to compare the mean CIP concentration in heated versus unheated bone.An ordinary one-way analysis of variance (ANOVA), followed by Fisher's least significant difference test, was conducted to compare the CIP concentration in bone versus the adjoining muscle.Treatment groups were compared for differences in mean CFU by two-way ANOVA followed by Tukey's multiple comparison test.All p values were two-sided, and p < .05 was considered statistically significant.

CIP-LTSL showed efficient encapsulation and temperature-dependent CIP release
The hydrodynamic diameter of CIP-LTSLs measured by DLS was in the range of 183.5 nm ± 1.91 (n ¼ 5), with a polydispersity index of 0.134 ± 0.009.CIP-LTSLs demonstrated a negative surface charge, as indicated by the zeta potential of À21.56 mV ± 0.58 (Figure 2(B)).TEM images confirmed the size distribution and showed CIP-LTSLs were spherical in shape (Figure 2(C)).Size analysis of CIP-LTSLs, stored at 4 C, over the period of 14 days, revealed that LTSLs were stable for eight days, after which the PDI of LTSL suspension started to increase (Figure 2(D)).CIP loading via transmembrane ammonium sulfate gradient yielded an encapsulation efficiency of >70% in the CIP-LTSLs.Percent CIP release from CIP-LTSLs started gradually at 39 C and was rapid and complete with 100% release at 40-42 C (Figure 2(E)).CIP-LTSLs stored in PBS at 4 C did not show any noticeable change in CIP release profile in response to temperature for 48 h (Figure 2(F)).

CIP-LTSLs induced anti-bacterial activity and efficient attachment to biofilms
MRSA cultures treated with CIP or CIP-LTSLs in combination with HIFU showed significantly higher bacterial killing ($60% higher) compared to untreated control, blank LTSLs ± HIFU and was similar to free CIP treatment (positive control) (Figure 6(A)).This was also confirmed by MRSA bioluminescence in treated groups (Figure 6(B)).In addition, Dio-LTSLs demonstrated efficient binding to MRSA biofilms within 24 h (Figure 6(C)).

Implant associated MRSA bone infection was confirmed in radiological and histological examinations
All rats survived the metal-implant surgery.Post-operation, the infected rats showed an initial decrease in body weight and mild limping.These signs gradually disappeared within 3-7 days post-surgery, and most rats exhibited full weightbearing gait with no signs of distress or pain.No marked differences in rectal temperatures were found in the rats postoperatively.
Analysis of radiographs at 10 days post-infection showed signs of osteomyelitis with a clear difference between the infected right femur versus the uninfected left femur in the rats.These signs included periosteal reaction and osteolysis of the femur as well as septic arthritic changes at the knee joint (Figure 4(A)).This was also confirmed by histopathological analysis of infected bone, in which salient features of osteomyelitis, including but not limited to loss of periosteal lining of bone and presence of extensive inflammation, fibrosis and clumps of bacteria, especially around the distal end of femur were noted (Figure 4(B)).Specifically, bone necrosis and subperiosteal scalloping of cortex were observed, along with the presence of intracellular bacteria and the infiltration of immune cells in the infected regions.

Combining HIFU thermal therapy with CIP-LTSLs increased bone CIP concentration
Treatment with HIFU caused mild discomfort in the rats immediately post-treatment, but it was alleviated within 6 h post-treatment, without any significant impact on ambulation (Figure S1).CIP-LTSL alone treatment delivered 1.33 ± 0.71 mg CIP/g to the bones but adding HIFU increased CIP levels to 2.16 ± 0.07.HIFU combined with free CIP also achieved 1.47 ± 0.16 mg CIP/g of tissue.Overall, CIP-LTSL þ HIFU was more efficient than CIP or CIP-LTSL alone in drug delivery.The targeted CIP delivery to the heated bone regions compared to the adjacent muscle (bone-to-muscle CIP ratio) was also found to be 1.4-fold and 2.2-fold greater for the CIP-LTSL þ HIFU group compared to the free CIP ± HIFU and CIP-LTSL groups, respectively (p < .05; Figure 5(B)).

Combinatorial therapy enhanced MRSA biofilm eradication
The tissue CFU counts were analyzed for heated bones, adjoining muscles (the lateral muscle layer directly overlying the heated bone region and the medial muscle farther away from the HIFU heated regions), skin and joint space in all treatment groups.In line with CIP delivery, the treatment efficacy of the CIP-LTSL þ HIFU group was best observed in the heated bones, which showed a 2.3-log and 2.9-log decrease in tissue MRSA load compared to the HIFU alone and free CIP þ HIFU groups, respectively (p < .05,p < .005; Figure 5(D)).CIP-LTSL þ HIFU treatment also decreased the bacterial load by 1.4-log and 1.1-log in the overlying lateral muscles and medial muscles, respectively, compared to the control (n ¼ 5; data not shown).This trend was also noted in the skin and joints compared to the other treatment groups.
To verify the bactericidal effects of the treatment, the bacterial burden was visualized by SEM imaging of the wire explants (Figure 7).Untreated control wires showed a thick film of S. aureus with extracellular matrix.Treatment with HIFU and CIP alone induced changes in the density and structure of the S. aureus biofilm compared to the control.However, a greater reduction of the extracellular matrix network was observed with the CIP-LTSL treatment.Furthermore, adding HIFU to CIP-LTSL treatment achieved the greatest reduction in the overall bacterial density and biofilm structure, as evidenced by reduced extracellular matrix cover and the presence of sparsely scattered S. aureus cocci.These results were verified in the H&E histopathological analysis of the bones and surrounding muscle layer, which exhibited a reduction in the bacterial burden and extent of tissue necrosis in the treatment groups compared to the non-treated control group (Figure S2).Overall, the CIP-LTSL þ HIFU group consistently decreased the S. aureus burden and the extent of bone tissue necrosis compared to the other HIFU treated groups (Figure 8).

Discussion
Eradication of biofilms from implant materials is a challenging aspect of orthopedic surgery [3].In particular, S. aureus biofilms that extend to bones and intramedullary tissues are a serious complication of open fractures and surgical repairs [13].These biofilms show poor antibiotic penetration, making chemotherapy challenging [14,15].To address this, HIFU has been investigated as a means to improve biofilm therapy in vitro and in vivo [16][17][18][19].Rediske et al. combined continuous ultrasound at 100 mW/cm 2 with gentamicin, but this approach did not reduce bacterial viability.In contrast, at 300 mW/cm 2 , the bacterial killing was significant, but it was also associated with some skin damage [19].Rieck et al. measured the killing of MRSA murine abscesses with moderate HIFU temperature (MT: 52.3 C ± 5.1 C) and high temperature (HT: 63.8 C ± 7.5 C).The MT and HT groups reached target temperatures after four 9-s ultrasound exposures applied in a square 1 Â 1 mm grid, with a 1 min pause between exposures.This resulted in a significant reduction in MRSA bacterial count in the treated areas compared with the untreated control at days 1 and 4 post-treatment [20].To further augment the HIFU effects with antibiotics, we innovated by combining localized HIFU hyperthermia with novel CIP-LTSLs to show improved outcomes against S. aureus-induced murine abscesses [21].We found that our combinatorial approach improved CIP delivery and also enhanced S. aureus clearance compared to the unheated control [21].Unlike superficial abscess wounds, delivering antibiotics through the calcified tissues of bone is highly challenging due to ultrasound attenuation of heat.The objective of this study was to determine whether combining HIFU with CIP-LTSLs is similarly capable of effectively clearing S. aureus bone biofilms via heat-targeted and mechanical effects of sonic energy.
Several animal models that mimic the presentation and pathophysiology of osteomyelitis have been employed in previous research [20,[22][23][24].In this study, we utilized a rat model of osteomyelitis.A sterile orthopedic K-wire was surgically inserted into the distal end of rat femurs to mimic intramedullary nailing.To be as close as possible to clinical disease representation, no promoters of infection were used besides the implant and S. aureus.We used S. aureus because it is the most common pathogen in osteomyelitis infections [3,25], and over 50% of clinical cases are caused by methicillin resistant strains [26,27].The number of bacteria inoculated in previous studies ranged from 10 1 to 10 9 CFU of S. aureus [24,[28][29][30][31][32][33].In our experiment, we used an inoculum of 10 6 CFU of a MRSA strain injected directly into the medullary cavity of the rat femur.Histopathological and radiological evaluation showed clinically classifiable signs of osteomyelitis infection after 10 days (Figure 4), including acute, destructive, chronic, localized and stable infection in all animals with the selected dose of MRSA.
Over the years, the development and spread of multiple antibiotic-resistant organisms have gained much attention [27].Soft tissue and bone can be invaded by MRSA, S. epidermidis, Pseudomonas aeruginosa, Klebsiella pneumoniae and other pathogens [34,35].CIP is a fluoroquinolone antibiotic that is active against both Gram-negative and Gram-positive bacteria.It is indicated for the treatment of a variety of bacterial infections, including bone and joint infections [36].Although resistance to fluoroquinolones, including CIP, can emerge rapidly in MRSA and P. aeruginosa spp.infection [37], Mulligan et al. recently reported that CIP can still be used for the eradication of MRSA colonization in combination therapies [38].Therefore, our motivation to use CIP in this study was multi-factorial.CIP is a broad-spectrum antibiotic, and the possibility that pathogens will develop resistance against it allowed us to assess whether HIFU can reverse antibiotic resistance, thereby improving outcomes against drug-resistant bacteria.HIFU also enhances the effectiveness of antibiotics by promoting their uptake by bacteria [39].As HIFU is a rapid focal technique, the bacteria may not have time to adapt to the applied stresses [40].All these factors motivated us to incorporate CIP into an LTSL regimen combined with HIFU.
In this experiment, we chose to use an LTSL-based delivery system based on our prior developed approach, and also similar findings from other research groups [41,42].The excellent biocompatibility of LTSLs and their controlled release of loaded antibiotics and biofilm penetrating features are clinically relevant [43].Ferreira et al. tested positively charged liposome attachment and their anti-biofilm activity against S. aureus biofilms.While strong attachment to biofilms was noted, it did not translate into improved anti-biofilm efficacy.In contrast, negatively charged liposomes showed greater therapeutic potential [44].Our LTSLs have a slightly negative zeta potential and have been widely used for localized delivery of anti-cancer drugs (e.g., doxorubicin) [45][46][47][48].We found that HIFU hyperthermia combined with CIP-LTSLs similarly achieved higher bacterial killing in vitro (Figure 6(A)).It may be noted that in in vitro experiments, CIP-LTSLs alone showed bacterial killing at 37 C compared to untreated control, which was further enhanced significantly with co-treatment of bacterial cultures with HIFU (p < .005).This is not unexpected as prolonged incubation (18 h) of LTSLs at 37 C can induce drug release possibly because of the physical strain of constant shaking conditions (200 rpm) and active interaction of lipid bilayer of LTSLs with media component and bacterial cells.Similar observations were seen by our group [7] and others [49,50].In contrast, HIFU plus CIP-LTSLs achieved improved CIP delivery and efficacy compared to the CIP-LTSL alone and free CIP þ HIFU groups in vivo (Figure 5(B,D)).Likely, the increased bone perfusion mediated by stable and long duration hyperthermia by HIFU induced greater LTSL uptake in the infected tissues [47,51,52].We also theorize that an increase in localized temperature by HIFU resulted in increased metabolic activity in bacterial cells, which in turn facilitated the uptake of CIP by the biofilm bacteria.In support of this premise, Fleury et al. studied transcriptomic and metabolic data from S. aureus exposed to a sub-lethal (43 C) temperature and found that metabolic activities, including ATP-generating pathways, were activated [53].
Treatment of implant-associated osteomyelitis consists of implant removal, extensive surgical debridement and prolonged antibiotic treatment.Additionally, implant coatings to inhibit bacterial adhesion with antibiotics, antiseptics or metal ions can be attempted [3,28].However, to the best of our knowledge, the application of noninvasive means to improve the therapeutic efficacy of HIFU for MRSA biofilm treatment in an animal model of implant-associated osteomyelitis has not been investigated prior to this study.Herein, we show a correlation of decreased S. aureus load in the heated bones as well as adjacent and overlying muscles with improved CIP-LTSL delivery using our combination therapy (Figure 5).Our SEM images also showed a distinct reduction of biofilm biomass on the wires explanted from the rat femurs treated with CIP-LTSL þ HIFU (Figure 7).This is similar to our previous findings of damaged S. aureus cells in abscesses exposed to 42-46 C using HIFU [21].Gera and Doores similarly found that E. coli treated with ultrasound had damaged cell walls and cell membranes [54].Our results agree with previous in vitro studies demonstrating biofilm detachment upon mild hyperthermia treatments [55,56].Our visualization of HIFU treatment effects on biofilms on explanted intramedullary wires provides important information about the effects of this therapy, and it indicates that this approach could further improve outcomes of an antibiotic in vivo.Interestingly, our histopathology analysis suggested that all treatment groups exhibited reduced bacterial burden and tissue necrosis compared to the untreated control group (Figure 8), without significant differences among the different treatment groups.This discrepancy between histopathology and SEM results could be due to the limitation posed by the tissue collection process.The femurs had to be sawed open to remove the intramedullary wire without disturbing the biofilm biomass.As infected bones are fragile, some bones were fragmented during this process that may have influenced histological enumerations.Finally, some treated rats experienced mild-to-moderate pain with HIFU exposures.This could be managed medically using anti-inflammatory drugs.Thus, we propose that the treatments did not have a significant impact on ambulatory status with the chosen ultrasound parameters.In summary, we demonstrated for the first time the feasibility of utilizing HIFU for physical eradication of MRSA biofilm bone infections or, at a minimum, biologically significant reduction of bacterial biomass with CIP-LTSLs in a non-surgical manner in rat bones.While the results from the rat studies are promising, further studies are needed to evaluate the safety and efficacy of this approach in humans.In particular, evaluation of the optimal dosing and administration of CIP-LTSL and HIFU in humans, as well as the long-term safety and effectiveness of the approach should be conducted.Although the current study assessed the effects of two treatments, complete eradication of biofilms in humans may require additional optimization of treatment duration, antibiotic combinations and frequency.Overall, the use of CIP-LTSL and HIFU for the treatment of bone infections represents an innovative and potentially transformative approach for treating a complex and difficult-to-treat clinical problem in animals and humans.With further research and development, this approach may ultimately offer significant benefits to patients suffering from bone infections.

Figure 1 .
Figure 1.Graphical representation of minimally invasive HIFU hyperthermia (40-45 C) combined antibiotic chemotherapy approach for the treatment of implant biofilm-associated osteomyelitis.Created with BioRender.com.

Figure 2 .
Figure 2. Ciprofloxacin-loaded low temperature sensitive liposomes (CIP-LTSLs) released the payload when heated to $40 C or greater.(A) Graphical representation of CIP-LTSL synthesis (created with BioRender.com).Formulated CIP-LTSLs showed homogenous size distribution with negative zeta-potential (B), were spherical in shape (C) and prolonged stability for eight days stored at 4 C determined by measuring changes in NP size and PDI (D).(E) CIP-LTSLs showed a narrow drug release range of 39-41 C and released 100% of entrapped CIP at 41 C. (F) CIP-LTSLs stored in PBS at 4 C were stable and demonstrated similar CIP release profiles from 0 to 48 h.

Figure 3 .
Figure 3. Experimental setup for HIFU hyperthermia treatment.(A) A 1.5 MHz HIFU transducer with a coaxially aligned imaging probe was used for targeting and treatment guidance.Anesthetized rat (yellow arrow) in a holder was mounted on a computer-controlled 3D positioning system and lowered into a water bath filled with circulating degassed water maintained at 37 C. (B) HIFU hyperthermia was achieved by selecting a region of interest with a raster treatment pattern at the bone-muscle interface while delivering the specified treatment protocol at each focal point (created with BioRender.com).

Figure 4 .
Figure 4. Establishment of the rat model of MRSA biofilm implant-associated osteomyelitis.(A) Radiograph at 10 days post-infection showed signs of periosteal reaction, osteolysis and swelling in the infected right femur compared to the uninfected left femur.(B) Histopathological changes in the rat model of osteomyelitis in the femur illustrated salient features of osteomyelitis: (i) extensive inflammation and fibrosis in the periosteal lining of the bone; (ii) pieces of necrotic bone (black arrow) surrounded by micro-abscess, inflammation and fibrosis; (iii) inflammation and presence of clumps of bacteria (black arrowheads) in the distal end of the femur; (iv) presence of immune cells with intracellular bacteria (black arrows).Scale bars ¼ 100 mm.

Figure 5 .
Figure 5. CIP-LTSL þ HIFU treatments significantly increased antibiotic delivery to treated bones and reduced local bacterial load.(A) Rats were infected on day 0, treated with CIP-LTSLs or free CIP ± HIFU on day 11, and sacrificed 24 h later, and then tissue CIP concentration was estimated (created with BioRender.com).(B) A significantly high targeted CIP delivery in heated bones compared to adjoining muscle in CIP-LTSL þ HIFU group compared to other groups was observed.(C) Experimental timeline of combinatorial treatment in rats with established infections.Rats were infected on day 0, treated with CIP-LTSL or free CIP ± HIFU on days 11 and 14, and sacrificed on day 15.Tissues were processed for bacteriological and histological analyses (created with BioRender.com).(D) HIFU-mediated targeted elimination of MRSA with CIP-LTSLs was seen in CIP-LTSL þ HIFU group.Log 10 CFU per ml of homogenized HIFU targeted bones between different treatment groups showed a significant decrease in CFU in the combination group compared to untreated control and monotherapies.Significance was calculated using oneway ANOVA with Fisher's least significant difference test.Ã p < .05,ÃÃ p < .005.

Figure 6 .
Figure 6.Combination of CIP-LTSL and HIFU achieved the greatest MRSA killing in bacterial culture in vitro.Significant decreases in MRSA numbers estimated as CFU (A) and bioluminescence (B) were observed in cultures treated with CIP-LTSL þ HIFU compared to other treatment groups, similar to CIP-alone treated positive control.(C) Time-resolved imaging of fluorescent DiO-labeled LTSLs incubated with MRSA biofilm for 24 h indicated efficient biofilm binding.

Figure 7 .
Figure 7. Scanning electron microscopy (SEM) used to observe the effect of CIP-LTSL þ HIFU treatments on biofilm-contaminated wire explants taken from infected bones showed a reduction in bacterial burden relative to other groups.The analysis was conducted on a single specimen (n ¼ 1).