Mild magnetic hyperthermia is synergistic with an antibiotic treatment against dual species biofilms consisting of S. aureus and P. aeruginosa by enhancing metabolic activity

Abstract

The wound biofilm infections that develop tolerance to standard-of-care antimicrobial treatment has been increasing. The objective of this study was to demonstrate a proof-of-concept of mild magnetic nanoparticle (MNP)/alternating magnetic field (AMF) hyperthermia as an anti-biofilm therapy against multispecies biofilm infections. Here, using both an in vitro cell culture and in vivo murine model of wound infection, we demonstrate that MNP/AMF hyperthermia applied at a mild thermal dosage was synergistically effective against dual species biofilm infection consisting of S. aureus and P. aeruginosa when combined with a broad-spectrum antibiotic, ciprofloxacin (CIP). For example, the combined treatment of MNP/AMF hyperthermia and CIP to the wounds of diabetic mice (db/db mice) significantly reduced the CFU number of S. aureus and P. aeruginosa by 2-log and 3-log, respectively, compared to the untreated control group, whereas either mild MNP/AMF hyperthermia or CIP treatment alone had little effect on the eradication of both bacteria. Our gene microarray data obtained from the culture of S. aureus biofilm suggest that mild MNP/AMF could shift the expression of genes for cellular respiration from anaerobic fermentation to an aerobic glycolytic/tricarboxylic acid cycle (TCA) pathway, implicating that the beneficial effect of mild MNP/AMF hyperthermia on the increased susceptibility of biofilm bacteria to an antibiotic treatment is associated with an increased metabolic activity. Taken together, our results support the translational potential of mild MNP/AMF as an adjunctive therapy that can be combined with a broad-spectrum antibiotic treatment for the management of wound biofilm infections associated with multispecies bacteria.

Introduction

The wound biofilm infections that develop a tolerance to standard-of-care antimicrobial treatment has been increasing [1–7]. We previously reported magnetic nanoparticle (MNP) and alternating magnetic field (AMF)-induced hyperthermia as a non-pharmacological strategy to combat biofilm infections [8], in which highly localized heat was generated on the surface of MNPs that are delivered to bacterial pathogens upon the application of external AMF. In the study, we demonstrated the capacity of MNP/AMF hyperthermia to target and inhibit the growth of bacterial pathogen using both in vitro and in vivo models of Staphylococcus aureus (S. aureus) biofilm infection. However, the complete eradication of bacterial pathogens required the application of MNP/AMF hyperthermia at a higher thermal dose, which could elicit non-specific thermal damage in the host tissue.

To address this, we recently demonstrated that MNP/AMF hyperthermia applied at a mild thermal dose, which is non-toxic to host mammalian cells, could be effective in enhancing the antibiofilm efficacy of conventional antibiotics in an in vitro model of S. aureus biofilm [9]. This finding is in line with previously published studies showing that general hyperthermia therapy could render biofilm bacteria to be susceptible to conventional antibiotics [10,11]. However, the detailed mechanism by which mild MNP/AMF hyperthermia sensitizes the antibiofilm effect of antibiotics has yet to be elucidated.

A critical challenge in the treatment of chronic wounds lies in the nature of biofilm infection formed by multispecies bacteria in a wound [7,12,13]. The common bacteria involved in chronic wounds include S. aureus, Enterococcus faecalis, Pseudomonas aeruginosa (P. aeruginosa), Klebsiella pneumoniae, and Acinetobacter baumanii [14–16]. The multispecies biofilms composed of those pathogens exhibited a higher level of antimicrobial tolerance than the biofilms formed by single species bacteria [17–19]. However, it has not been shown whether mild MNP/AMF hyperthermia could be effective against a wound biofilm infection associated with multispecies bacteria.

The goal of this study was to demonstrate a proof-of-concept of mild MNP/AMF hyperthermia in combination with a broad-spectrum antibiotic as an anti-biofilm therapy against multispecies biofilm infections. S. aureus and P. aeruginosa are the prevalent types of bacterial pathogens found in polymicrobial infections in chronic wounds including diabetic ulcers [5,20]. In this study, we employed a wound biofilm model composed of dual species of bacteria, S. aureus and P. aeruginosa, in diabetic mice. Additionally, in order to gain more insight into the mechanism by which mild MNP/AMF hyperthermia sensitizes the antibiofilm effect of antibiotics, we examined the effect of mild MNP/AMF hyperthermia on the global transcriptional profiling in the biofilm phase of bacteria by using a full-genome microarray analysis in S. aureus biofilm. Our results showed that the application of mild MNP/AMF to the wounds of diabetic mice infected with dual spices biofilm formed by S. aureus and P. aeruginosa synergistically enhanced the antimicrobial efficacy of ciprofloxacin (CIP), a broad-spectrum of antibiotics. Additionally, our full-genome microarray analysis revealed that mild MNP/AMF hyperthermia altered the transcriptional profiling of biofilm phase of bacteria, especially in the genes associated with cellular metabolism.

Materials and Methods

Preparations of S. aureus and P. aeruginosa

S. aureus (ATCC 6538 strain) and P. aeruginosa (PAO1, ATCC BAA-47) were obtained from the American Type Culture Collection (ATCC). Both bacterial strains were streaked onto the Tryptic Soy Agar (TSA) (BD Biosciences) and the streaked plates were incubated overnight at 37°C. A single colony of bacteria was inoculated to the 5 mL of Tryptic Soy Broth (TSB) for S. aureus and Luria-Bertani (LB) broth (BD Biosciences) for P. aeruginosa, respectively. They were then incubated under aerobic conditions at 37°C in a shaker at 180 rpm overnight until they reached a stationary phase. For the preparation of planktonic phase of S. aureus and P. aeruginosa, the cells were pelleted by centrifugation at 3,500g at 4 °C for 7 min and diluted with TSB and LB, respectively, to obtain the desired concentrations of bacteria (1×10³ - 1×10⁶ CFU/mL).

In vitro model of dual species biofilm formed by S. aureus and P. aeruginosa

The dual species biofilm formed by S. aureus and P. aeruginosa were established using a static model described previously with some modifications [21,22]. Bacteria suspensions of stationary phase growth of P. aeruginosa and S. aureus were diluted in sterile phosphate buffer solution (PBS) at final concentrations (1x10³ - 1x10⁶ CFU/mL) in each tube. The diluted solution was then mixed at a varying density between P. aeruginosa and S. aureus (1:1, 1:10, 1:100, and 1:1000) and 2 μL of the mixed suspension was inoculated on to a 6 mm diameter of polycarbonate membrane filter (0.2 μm pore size, GE). The filter was incubated on TSA plate and incubated at 37°C for 48 h. After the culture, for subsequent CFU counting on selective agar media, the biofilm matrix was disrupted through vortexing (at 900 rpm) followed by sonication in a sonication bath (Branson Ultrasonic bath) for 5 min. The Mannitol salt agar and Triclosan agar were used as a selective agent for S. aureus and P. aeruginosa, respectively (Figure S1A).

The susceptibility of single species and dual species biofilm to ciprofloxacin

The susceptibility of either single species or dual species biofilm to a ciprofloxacin treatment was determined using a broth macro-dilution method [23]. CIP is a broad-spectrum fluoroquinolone antibiotic used clinically to treat both S. aureus and P. aeruginosa infections. Either single species biofilm formed by S. aureus or P. aeruginosa alone or dual species biofilm formed by the co-culture of S. aureus or P. aeruginosa was prepared as described above. The biofilms formed on 6 mm diameter polycarbonate membrane filters were transferred to sterile 96-well flat-bottom polystyrene tissue culture plates (BD Falcon) and they were treated with CIP (Sigma-Aldrich) at concentrations ranging from 0.1 μg/mL to 3 mg/mL for 18 h at 37°C. The cells were then plated on Mannitol salt agar for S. aureus CFU counting and Triclosan agar for P. aeruginosa CFU counting.

Biofilm mass assay

The reduction of biofilm mass to a CIP treatment was conducted by crystal violet (CV) staining method. Briefly, 96-well polystyrene flat-bottomed microplate containing either single species or dual species biofilms were washed with PBS three times to remove any non-adherent bacteria. The biofilms were then fixed with 200 μL of 99% (v/v) methanol for 15 min and left to air-dry for 30 min in a laminar flow. The fixed biofilm cells were stained using filtered 0.5% crystal violet (Sigma-Aldrich) for 5 min at room temperature. The excess stain was removed by rinsing with running tap water, then air dried, and crystal violet bound cells were solubilized with 200 μL of 33% (v/v) of acetic acid (Merck). The biofilm mass was quantified through the measurement of optical density of the sample at 570 nm using a spectrophotometer (SpectraMax^® M4 Multi-Mode Microplate Reader, Molecular Devices).

Measurements of temperature increase (ΔT) during the exposure of MNP/AMF hyperthermia

The procedure for the application of MNP/AMF hyperthermia to the biofilm was performed as described in our previous study [8]. In brief, varying concentrations of MNPs (Ferumoxytol, AMAG) ranging from 1 to 3 mg/mL were added to the wells of S. aureus biofilm (in 200 μL solution) pre-formed on a polystyrene 8-well chamber slide and incubated for 2 h. For the application of AMF, the chamber slide was positioned to the water-cooled coil chamber of the AMF generator (5kW power operated at 300-450kHz, MSI Automation) and applied with AMF for 10 min at the field strength of 60 kA/m and frequency of 375 kHz. The real-time monitoring of temperature in the solution of biofilm during the AMF exposure was conducted using a fiber optic temperature probe (Neoptix) placed in the solution.

Determination of cumulative equivalent minutes at 43°C (CEM₄₃)

The CEM₄₃ was calculated from equation as described [24], ${\mathrm{CEM}}_{43}=\sum {\text{t}}_i\times {\text{R}}^{(43-T)}$, where T is the average temperature during i-th time interval t_i (min), and R is a factor to compensate for temperature change, which is set at 0.25 for T≤43°C and 0.5 for T>43°C. The average temperature (T) was determined from the plot of temperature vs time at a given time interval during the application of MNP/AMF hyperthermia.

Antibiotic susceptibility of dual species biofilm to MNP/AMF hyperthermia

The dual species biofilm containing S. aureus and P. aeruginosa was formed on an 8 well-chamber slide for 48 h. The biofilm was treated with MNPs (1 or 2 mg/mL, Ferumoxytol, AMAG) for 2 h at 37 °C and then applied with AMF for 10 min at the field strength of 60 kA/m and frequency of 375 kHz. Following the application of MNP/AMF, the biofilms were treated with CIP at a concentration of 50 μg/mL for 18 h at 37 °C and the CFU numbers of S. aureus and P. aeruginosa were counted as described above.

In vivo model of dual species biofilm infection in wounds of diabetic mice

Diabetic db/db mice (BKS.Cg-Dock7^{m +/+} Lepr^db/J, male mice at age of 8-12 weeks old) were obtained from Jackson laboratory (Bar Harbor, ME). Mice were anesthetized with 1.5% isoflurane gas inhalation and one full-thickness, circular wound was made on the dorsal surface of a mouse using a 6 mm sterile biopsy punch (Acuderm Inc.). The wound of each mouse was inoculated with 20 μL of bacteria solution mixed with P. aeruginosa (1x10³ CFU/mL) and S. aureus (1x10⁶ CFU/mL) and covered with a transparent, semipermeable Tegaderm dressing (3M). At day 2 post infection, either 20 μL of sterile saline, CIP (50 μg in 20 μL of sterile saline) only, MNPs (40 μg of Ferumoxytol in 20 μL saline) only, or CIP (50 μg) mixed with MNPs (40 μg of Ferumoxytol in 20 μL saline) was topically injected to the area of wound infection. The mice treated with MNPs were placed in a magnetic coil chamber and AMF was applied for 10 min at the frequency of 375 kHz and field strength of 60 kA/m. At day 3 post-infection, the wounded skin was excised and homogenized using a tissue homogenizer (OMNI tissue homogenizer, OMNI Inc.), and the homogenates were serially diluted and plated on the agar with selective agent for the CFU counting of S. aureus and P. aeruginosa. The CFU number in the wound was standardized with respect to the weight of excised wound. Three mice were used for each experimental group. The experimental protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Kent State University.

Scanning electron microscopy (SEM) imaging of wound biofilm

Wounded skins were harvested from db/db mouse at day 2 post-infection and fixed with 3% (v/v) glutaraldehyde for 10 min at room temperature, followed by overnight incubation at 4 °C. The samples were treated with a series of ethanol washes (30, 50, 70, 95 and 100%, 10 min for each concentration) and then subsequently placed in a critical point dryer to replace ethanol with CO₂. The dried samples were then mounted to an aluminum stub and sputter-coated with gold-palladium for 90 seconds and examined in a scanning electron microscope (Hitachi S-4700 SEM, Hitachi High Technologies America, Inc.) at x10,000 magnification.

Gene microarray analysis

The microarray analysis was run in the genomics shared resource-comprehensive cancer center in Ohio State University (Columbus, OH). RNA samples were processed according to the Affymetrix SensationPlus FFPE Amplification and 3′IVT Labeling Kit protocol and also as described previously [25]. Briefly, 50 ng of total RNA was reverse transcribed to produce sense RNA via in vitro transcription (IVT), followed by single-strand cDNA synthesis. The cDNA was fragmented, end-labelled and hybridized to the Affymetrix GeneChip S. aureus Genome Array (Affymetrix, Cat. No. 900514), which contained probe sets to over 3,300 S. aureus open reading frames. GeneChip arrays were scanned using an Affymetrix 3000 7G scanner. The resultant data (CHP files) were submitted to the Affymetrix^® Transcriptome Analysis Console (TAC) software for Gene Level Differential Expression Analysis. The gene expression analysis was performed from three independent mRNA samples for each group. Genes with a p value < 0.05 and threshold values of ≥ 2 and ≤ – 2-fold change between the MNP/AMF-treated S. aures biofilms and the untreated control was defined as significantly differentially expressed.

Determination of metabolic activity

S. aureus biofilm was prepared on 96-well strip plate (SPL Strip Immunoplate) and the metabolic activity of the biofilm in response to MNP/AMF hyperthermia was determined using a resazurin-based assay. In brief, either MNP/AMF-treated (at the thermal dose of CEM43=6 min) or untreated control S. aureus biofilm was rinsed and 120 μl of a resazurin solution (Promega) was added to each well. The plate was incubated for 20 min at 37 °C and fluorescence was measured (at excitation wavelength of 560 nm and emission wavelength of 590 nm) using spectrophotometer (SpectraMax^® M4 Multi-Mode Microplate Reader, Molecular Devices). The values were expressed after normalizing the fluorescence intensity with respect to the number of viable bacteria in the well.

Statistical Analysis

Statistical analysis was performed using GraphPad Prism version 8.0 software. A two-tailed unpaired t-test was used to determine statistical significance between two groups. Statistical tests among multiple groups were analyzed using one-way ANOVA followed by Turkey’s posttest for secondary analysis for comparison. For all analyses, p-values of less than 0.05 was considered to be statistically significant. Data were presented as mean ± standard deviation (SD). The in vitro studies were run with at least three biological replicates, and each biological replicate had three technical replicates.

Results

The susceptibility of dual species biofilm consisting of S. aureus and P. aeruginosa

Since S. aureus and P. aeruginosa are major pathogens found growing together in biofilms in chronic wounds including diabetic wounds [5,20], we first sought to establish an in vitro model of polymicrobial biofilms formed by the balanced growth of both organisms. Since P. aeruginosa was shown to outcompete the growth of S. aureus under co-culture condition with an equal inoculation density [21], we reasoned that co-inoculating with increasing densities of S. aureus over P. aeruginosa would result in the balanced growth of both bacteria in the biofilm. To ascertain this, P. aeruginosa and S. aureus were inoculated on to a membrane filter at varying density ratios ranging from 1:1, 1:10, 1:100, and 1:1000 and co-cultured to form a stable biofilm. For the biofilm formed at the 1:1 inoculation ratio between P. aeruginosa and S. aureus, the CFU number of P. aeruginosa was significantly higher than that of S. aureus by ~3 log (Figure S1B). Increasing the density of S. aureus could significantly increase the viable number of S. aureus in the biofilm, resulting in the difference between S. aureus and P. aeruginosa by 2-log at the 1:100 inoculation ratio (p<0.05) and by less than 1-log at 1:1,000 inoculation (p>0.05). Thus, in the subsequent studies, dual species biofilm was formed by seeding P. aeruginosa and S. aureus at 1:1,000 density.

By establishing the protocol for dual species biofilm with a balanced growth of S. aureus and P. aeruginosa, we next determined the susceptibility of dual species biofilm to an antibiotic. We chose CIP as an antibiotic to be tested since it exhibits a broad-spectrum antibacterial efficacy against both S. aureus and P. aeruginosa in the planktonic phase. The minimum inhibitory concentration (MIC) values of CIP against planktonic phase of S. aureus and P. aeruginosa were measured to be between 0.1 to 0.5 μg/mL. However, when each strain of bacteria formed a biofilm by either S. aureus or P. aeruginosa, the use of CIP at 100 μg/mL was required to achieve a growth inhibitory effect on biofilm bacteria, 200~1,000 times higher concentration of CIP than the ones for planktonic phase of bacteria (Figure 1A). Importantly, it was even more difficult to eradicate the biofilm when both bacteria co-existed by forming a dual species biofilm. This was also evidenced by the incomplete eradication of biofilm mass by CIP for dual species biofilm, compared to the biofilm formed by single species bacteria (Figure 1B). Our result is consistent with previous reports that showed a reduced susceptibility of multispecies biofilm to antibiotics [21,26], supporting the validity of our dual species biofilm model.

Figure 1. The susceptibility of dual species biofilm consisting of S. aureus and P. aeruginosa to ciprofloxacin in vitro.

(A) Log (CFU/mL) reduction of S. aureus and P. aeruginosa in single species biofilms (Single-S. aureus and Single-P. aeruginosa) and dual species biofilms (Dual-S. aureus and Dual-P. aeruginosa) following the treatment of increasing concentrations of ciprofloxacin. (B) Biomass reduction in single species biofilms (Single-S. aureus and Single-P. aeruginosa) and dual species biofilms (Dual-S. aureus + P. aeruginosa) following the treatment of increasing concentrations of ciprofloxacin. The dual biofilms data is analysis of S. aureus or P. aeruginosa from the dual biofilm. N=3 per group. *: p<0.05 and #: p<0.01.

Determination of CEM₄₃

Based on our recent study showing the synergistic efficacy of mild MNP/AMF hyperthermia with conventional antibiotics against S. aureus biofilm in vitro [9], we sought to determine whether the mild MNP/AMF hyperthermia therapy could also be effective against biofilm formed by multispecies bacteria. As a type of MNP for this study, we used Ferumoxytol (Feraheme, AMAG Pharmaceuticals), which is an FDA-approved superparamagnetic iron oxide nanoparticle (SPION) to treat anemia associated with chronic kidney disease. We first determined the hyperthermia characteristic of Ferumoxytol by measuring the temperature increase (ΔT) for varying concentrations of Ferumoxytol solutions at the fixed AMF strength (60 kA/m) and frequency (375 kHz) (Figure 2A). The application of MNP/AMF at 1 mg/mL of Ferumoxytol for 10 min resulted in ΔT of 5.3 °C and the ΔT value was further increased to 11.6 °C and 16.1 °C for 2 mg/mL and 3 mg/mL of Ferumoxytol, respectively (Figure 2B), confirming the heating capacity of Ferumoxytol-MNPs.

Figure 2. Effects of MNP/AMF hyperthermia on the CEM₄₃ thermal dose.

(A) An experimental setup for the measurment of temperature in MNP (Ferumoxytol) solution during the application of AMF (B) Increase of temperature (ΔT) during the application of AMF as a function of MNP concentration (1-3 mg/mL). N=3 per group. (C) Estimation of MNP/AMF-induced CEM₄₃ (min) as a function of MNP concentration (1-3 mg/mL). The dotted red line indicates the reported threshold of safe thermal dose for the application of hyperthermia to skin tissue²⁴. N=3 per group.

Since CEM₄₃ has been widely used as an accepted metric for thermal damage assessment in human tissues in response to various hyperthermia treatments, we next determined the concentration of Ferumoxytol that could be applied to skin tissue within the safety margin of thermal dose by correlating the measured values of ΔT with CEM₄₃. The calculated values of CEM₄₃ ranged from ~1 min for 1 mg/mL, 20 min for 2 mg/mL, and up to 7,450 min for 3 mg/mL of Ferumoxytol-MNPs (Figure 3C). Given the reported value of CEM₄₃ of 41 min as a threshold thermal dose for the application of hyperthermia to a skin tissue [24], treatment conditions of Ferumoxytol-MNPs up to 2 mg/mL for 10 min AMF, equivalent to CEM₄₃ of 20 min, was determined to be an acceptable range of thermal dose for skin wounding experiments.

Figure 3. Effects of MNP/AMF hyperthermia on the killing of S. aureus and P. aeruginosa in dual species biofilm in combination with ciprofloxacin for varying thermal dose.

(A) An experimental setup for the application of MNP/AMF hyperthermia to the culture of dual species biofilm consisting of S. aureus and P. aeruginosa. (B) The effect of MNP/AMF hyperthermia alone or in combination with ciprofloxacin (CIP, 50 μg/mL) for varying CEM₄₃ thermal doses (CEM₄₃=1 and 20 min) on the log reduction of S. aureus and P. aeruginosa numbers (CFU/mL) in dual species biofilm. N=3 per group. *: p<0.01.

Mild MNP/AMF hyperthermia synergistically enhanced the susceptibility of dual species biofilms to ciprofloxacin in vitro.

By observing an antibiotic tolerance of dual species biofilm formed by S. aureus and P. aeruginosa to CIP, we next examined whether mild MNP/AMF hyperthermia would enhance the antibacterial efficacy of CIP against the biofilm. For this, the culture of dual species biofilm by S. aureus and P. aeruginosa was pre-treated with MNP/AMF at the thermal dose of CEM₄₃=1 min and CEM₄₃=20 min (Figure 3A) in the absence and presence of CIP (50 μg/mL). The CIP alone had little effect on the killing of both S. aureus and P. aeruginosa when they were grown together in the biofilm phase (Figure 3B). The application of MNP/AMF alone at the CEM₄₃ thermal dosage of 1 min had also little effect on the killing of both bacteria, while the combined treatment of MNP/AMF and CIP resulted in about 1-log reduction of S. aureus and 0.6-log reduction of P. aeruginosa in the biofilm. Increasing the level of CEM₄₃ thermal dose to 20 min resulted in a 2-log reduction of S. aureus and 1.4-log reduction of P. aeruginosa bacteria in the presence of CIP, while the application of MNP/AMF alone resulted in only a 0.6-log and 0.3-log reduction of S. aureus and P. aeruginosa, respectively (Figure 3B). Taken together, these results suggest that the application of mild MNP/AMF hyperthermia is synergistic with CIP against dual species biofilm.

In vivo validation of mild MNP/AMF hyperthermia in wound of diabetic mice infected by S. aureus and P. aeruginosa.

By observing the synergistic efficacy mild MNP/AMF with CIP against dual species biofilm consisting of S. aureus and P. aeruginosa, we next engaged in a study to validate its efficacy using an in vivo model by employing a murine model of wound infection in type 2 diabetic mice (db/db mice). Mice were randomly divided into 4 groups based on treatment conditions, which include untreated control, MNP/AMF only, CIP only, and MNP/AMF+CIP. Our in vitro protocol for inoculating a higher density of S. aureus over P. aeruginosa could establish the balanced growth of S. aureus and P. aeruginosa in the biofilm (Figure 1B). Thus, in this study, db/db mice were inoculated with the mixture of P. aeruginosa and S. aureus prepared at 1:1,000 ratio (1×10³ CFU of P. aeruginosa to 1×10⁶ CFU of S. aureus) onto the wounds at day 0 (Figure S2A). As observed in the in vitro culture experiment, the growth rate of P. aeruginosa was greater than S. aureus in wounds of db/db mice and the number of both bacteria in the wounds reached at a similar portion at day 2 post-infection (Figure S2B). The formation of dual species biofilm by S. aureus and P. aeruginosa in the wound was confirmed by SEM imaging of wound samples (Figure 4A).

Figure 4. The in vivo efficacy of combined treatment of mild MNP/AMF hyperthermia with ciprofloxacin in wounds of db/db mice infected with dual species biofilm.

(A) An experimental procedure for the application of mild MNP/AMF and CIP to the wounds of db/db mice infected with dual species biofilm consisting of S. aureus and P. aeruginosa. The SEM image shows the presence of S. aureus (spherical shape, white arrow) and P. aeruginosa (rod shape, yellow arrow) in the biofilm collected from wounds of db/db mice day 2 post-infection. (B) Temperature increases in wound surface and rectum of db/db mice measured during the application of MNP/AMF measured by fiber optic temperature probe. (C) The effect of combined mild MNP/AMF hyperthermia and CIP treatment on the reduction in CFU numbers of S. aureus and P. aeruginosa in wounds of db/db mice at day 3 post-infection. N=3 per group. *: p<0.05 and **: p<0.01.

At day 2 post-infection, the wounds of db/db mice were topically applied with either a solution of Ferumoxytol-MNPs (40 μg in 20 μL saline, equivalent to 2 mg/mL MNPs) only, CIP (50 μg per wound) only, or a mixture of Ferumoxytol-MNPs and CIP (40 μg MNPs+50 μg CIP in 20 μL saline) and the mice were subsequently applied with an AMF (375 kHz and 60 kA/m) for 10 min (Figure 4A). The application of MNP/AMF resulted in the increase of wound surface temperature by 4°C over the duration of 10 min (Figure 4B). Since the application of high frequency AMF has been associated with the generation of eddy currents that leads to non-specific body heating, changes in rectal temperatures were monitored as a measure of systemic temperature changes during the application of AMF. The increase of rectal temperature over 10 min by the application of AMF was measured to be about 1°C, suggesting that the application of mild MNP/AMF did not cause an adverse increase of systemic temperature. The CFU numbers of S. aureus and P. aeruginosa were quantified from the wounded skin harvested after 24 h post-MNP/AMF application (i.e., day 3 post-infection). The treatment of CIP alone reduced the CFU number of S. aureus and P. aeruginosa in wounds by 0.5-log and 1.5-log, respectively, compared to the untreated control group, while the application of MNP/AMF alone slightly reduced P. aeruginosa by 0.4-log but had little effect on the viability of S. aureus. (Figure 4C). The combined treatment of MNP/AMF and CIP to the wounds of mice significantly reduced the CFU number of S. aureus and P. aeruginosa by 2-log and 3-log, respectively, compared to untreated control group (Figure 4C).

The application of mild MNP/AMF hyperthermia altered the global transcriptional profiling in S. aureus biofilm.

Our results clearly support the beneficial effect of mild MNP/AMF hyperthermia on the eradication of biofilm when combined with conventional antibiotics. However, the detailed mechanism by which MNP/AMF hyperthermia sensitizes the antibiofilm effect of antibiotics has yet to be elucidated. As a first step to address this question, we examined the impact of MNP/AMF on the global transcriptional profiling of biofilm bacteria by using a full-genome microarray analysis in an in vitro culture of S. aureus biofilm. The microarray analysis revealed 170 genes whose expressions were significantly altered (fold change ≤ −2 and ≥ 2, p < 0.05) following the application of the MNP/AMF hyperthermia at a mild thermal dose (CEM₄₃=6 min). Among them, 74 genes were up-regulated while 96 genes were down-regulated (Figure 5A). The genes were further categorized into functional clusters, which include genes associated with cell wall remodeling, oxidative stress, efflux pump, stress responses, DNA repair, protein synthesis, heat shock protein (HSP), metabolic pathways and hypothetical proteins (Supplementary Table 1).

Figure 5. The effects of mild MNP/AMF hyperthermia on the metabolism of S. aureus biofilm.

(A) Volcano plot showing the gene level analysis to select differentially expressed genes after treatment with mild MNP/AMF (CEM₄₃=6 min). The data for all genes are plotted as fold change versus −log10 of the adjusted p-value. Green, gray and red correspond to genes with < −2-fold, −2 to 2-fold and > 2-fold differential expression, respectively. (B) Changes in the expression of the genes for anaerobic fermentation (ldhA, adhE, pflA and pflB) and genes for aerobic glycolysis/TCA pathway (Pyr, pckA, fda, and odhA) in response to mild MNP/AMF hyperthermia (CEM₄₃=6 min) in S. aureus biofilm. The values listed in the figures are the calculated means of fold change from three parallel microarrays. (C) Changes in metabolic activity in S. aureus biofilm in response to MNP/AMF (CEM₄₃=6 min) quantified by resazurin assay. N=8 per group. *: p<0.05.

Among the set of genes differentially expressed to mild MNP/AMF in S. aureus biofilm, we particularly focused on functional clusters that were either robustly increased or decreased. The genes for cell wall stress stimulon and heat-shock proteins were observed to be increased to mild a MNP/AMF. For example, the expression of genes for cell wall stress stimulon including murA, msrA1, tagH, bioD, lytH, and yidC were upregulated by 3 to 4-fold following MNP/AMF compared to an untreated control group. The expression of HSP related genes (ctsR, mcsA, mcsB, grpE, clpL, hrcA, and ctsR) were also upregulated by 4 to 11-fold (Supplementary Table 1), which is in line with previous studies of heat-shock responses in S. aureus [27]. In contrast, the expression of genes associated with protein synthesis and cell division were downregulated following the application of MNP/AMF, albeit modestly (2 to 4-fold). The transcription of ribosomal protein genes including, rplX, rplE, rpsN, rpsH, rplF, and rplR, was shown to be downregulated in S. aureus biofilm, compared to its planktonic phase [28]. Our result showed that the application of a mild MNP/AMF could increase the expression of those genes by 2-fold, compared to an untreated biofilm control group (Supplementary Table 1).

Of particular observation from the gene microarray data was the shift in the expression of metabolism-related genes. For example, the expression of genes for fermentation such as ldhA, pflA and pflB were significantly downregulated by 8- to10-fold following the application of MNP/AMF hyperthermia (Figure 5B). The ldhA gene is involved in the synthesis of fermentation enzymes like lactate dehydrogenases. The pflB and pflA genes play a role in acetate and ethanol formation, which were shown to be upregulated under anoxic conditions within biofilms [29]. In contrast, the expression of genes for glycolytic and tricarboxylic acid cycle (TCA) pathways such as Pyr, fda, odhA were significantly upregulated by 4 to 10-fold following the application of MNP/AMF hyperthermia (Figure 5B). The Pyr and fda are genes for key enzymes in the glycolysis pathway under aerobic respiration [30] and odhA is a gene for the TCA cycle [31].

The application of mild MNP/AMF hyperthermia was associated with an increased metabolic activity in S. aureus biofilm

Our gene microarray results imply that the MNP/AMF hyperthermia might shift cellular respiration of the biofilm phase of bacteria from an anaerobic fermentation to an aerobic glycolytic/TCA pathway. In our recent study, we demonstrated that the synergistic effect of mild MNP/AMF on the antibiotic efficacy in S. aureus biofilm was associated with an increased uptake of antibiotics [9]. It has been shown that the bacteria within biofilms exhibited reduced metabolic activity due to the limited oxygen and nutrients, and this has been associated with a low susceptibility to antibiotics [32–34]. These findings prompted us to examine whether the altered expression of genes for cellular respiration was linked to a change in metabolic activity in S. aureus biofilm following MNP/AMF. For this, the metabolic activity of S. aureus biofilm in the absence and presence of mild MNP/AMF hyperthermia (CEM₄₃=6 min) was assessed by resazurin assay, which has been used to assess the metabolic activity of bacteria residing in biofilms [35,36]. The metabolic activity of S. aureus was significantly increased for biofilm treated with MNP/AMF hyperthermia compared to the untreated control group of biofilms at both 1 and 4 h post-MNP/AMF (P< 0.05) and the activity reverted back to baseline at 18 h (Figure 5C).

Discussion

We recently reported that the application of mild MNP/AMF hyperthermia could be synergistic with antibiotics in eradicating single species biofilm formed by S. aureus [9]. Here, using both in vitro and in vivo murine models of biofilm infection, we demonstrate that mild MNP/AMF hyperthermia is also effective against dual species biofilm infection consisting of S. aureus and P. aeruginosa when combined with a broad-spectrum antibiotic, which alone became ineffective otherwise. Our protocol for inoculating the high density of S. aureus over P. aeruginosa enabled the formation of dual species biofilm with a balanced growth of both bacteria. CIP is a broad-spectrum antibiotic drug that has shown to be highly effective for both S. aureus and P. aeruginosa with low MIC values [37]. The dual species biofilm exhibited a higher tolerance to CIP than single species biofilm formed by either S. aureus or P. aeruginosa alone. In the current study, we employed a murine model of dual species biofilm in wounds of db/db mice, a genetic mouse model of type 2 diabetes associated with impaired wound healing and persistent infection [38,39]. In consistent with in vitro results, our in vivo study validated that the combined treatment of mild MNP/AMF and CIP could synergistically reduce the number of viable bacteria of both S. aureus and P. aeruginosa in wounds of the mice, compared to either CIP or MNP/AMF treatment alone.

What is the potential mechanism by which mild MNP/AMF hyperthermia can sensitize biofilm bacteria to be susceptible to antibiotics? We speculate that this is associated with heat shock-induced alteration of the metabolic state of bacteria in the biofilm phase. This is supported by our result showing that the S. aureus biofilm treated with mild MNP/AMF hyperthermia was sufficient to increase metabolic activity in the biofilm. We recently demonstrated that MNP/AMF-induced heat shock to the culture of S. aureus biofilm could significantly enhance the uptake of antibiotic to the biofilm phase of bacteria, with the response that is not dependent on antibiotic penetration through the biofilm matrix [9]. In an earlier study for testing P. aeruginosa biofilm susceptibility to ciprofloxacin, oxygen limitation and the resulting low metabolic activity, not the limited antibiotic penetration, was attributed to antibiotic tolerance in the biofilm [34]. The bacterial cells embedded in biofilms exhibit a suppressed metabolic activity and as such, an altered metabolic state can influence antibiotic susceptibility by altering antibiotic uptake [32,40]. It is interesting to note that, for gram-negative swarming bacteria including Proteus mirabilis, hyperthermia-induced enhancement of antibiotic susceptibility was observed to be through the alteration of a cell wall structure that resulted in the inhibition of penicillin-binding protein-2 [41].

Our gene microarray data suggest that mild MNP/AMF can shift the expression of genes for cellular respiration from anaerobic fermentation to aerobic glycolytic/TCA pathways in S. aureus biofilm. We speculate that this might contribute to an increased antibiotic uptake by bacteria in a biofilm phase. This speculation is supported by previous studies for targeting TCA metabolic pathways as an antimicrobial strategy. For example, the metabolites from the TCA cycle could improve the susceptibility of resistant bacteria to bactericidal antibiotics [42,43] whereas alterations in metabolism involving a decreased TCA cycle activity was associated with the tolerance of S. aureus to antibiotics including daptomycin [44]. The role of TCA cycle activity in antibiotic tolerance was also evidenced in a recent study using a polymicrobial biofilm model [19], in which decreased metabolic activity associated with decreased TCA cycle activity was found to be a mechanism for increased antibiotic tolerance within polymicrobial cultures formed by S. aureus and Candida albicans, compared to monocultures of S. aureus. Future study should be directed towards determining whether a shift in metabolic activity is functionally linked to the beneficial effect of mild MNP/AMF hyperthermia against biofilm infections.

This study has some limitations. Firstly, in this study, we used standard strains of bacteria and it is necessary to assess the therapeutic efficacy of MNP/AMF hyperthermia using clinical bacterial isolates for future clinical translation. Additionally, the dose of CIP used in this study (50 μg/mL) for combined therapy with MNP/AMF hyperthermia is relatively high for systemic treatment and it needs to be further assessed using standard dose of antibiotics. Secondly, future study should be directed towards optimizing treatment conditions such as a therapeutic window and frequency of MNP/AMF application and examining long-term effects of the therapy. A therapeutic window of MNP/AMF treatment on the biofilm needs to be carefully determined whether simultaneous or post exposure of antibiotic would be beneficial in the combination therapy. Additionally, our result of metabolic activity showed that single session treatment of MNP/AMF hyperthermia may not be sufficient to induce a sustained increase in metabolic activity. It would be interesting to further examine whether repeated application at a lower thermal dose would be effective in stimulating a sustained metabolic activity and thereby suppressing the potential regrowth of biofilm.

In summary, our results support the translational potential of mild MNP/AMF as an adjunctive therapy that can be combined with a conventional antibiotic treatment for the management of wound biofilm infections associated with multispecies bacteria. Additionally, our findings imply that the beneficial effect of mild MNP/AMF hyperthermia on the increased susceptibility of biofilm to an antibiotic treatment is associated with an increased metabolic activity, which in turn facilitates an increased antibiotic uptake by bacteria in biofilm phase.

Data availability statement

The data that support the findings of this study are available from the corresponding author (MK) upon reasonable request.