Fortified interpenetrating polymers – bacteria resistant coatings for medical devices

Nanocapsule-mediated eugenol release from an interpenetrating polymer network coating reduces bacterial binding on medical devices.


Table S2
Relative fluorescence intensity of the supernatant after 20, 30 and 60 h incubation of monoliths (with fluorescent nanocapsules) in water.

Optimisation of surfactant composition for emulsification
Nanoemulsions containing 5% clove oil and 10% lauryl acrylate were prepared with HPLC grade water using the phase inversion temperature (PIT) method. The total concentration of the surfactant was maintained at 10% and various ratios (1:0, 3:1, 1:1, 1:3 and 0:1) of Span ® 80 and Kolliphor ® RH40 (hydrophilic-lipophilic balance values of 4 and 12 to 14, respectively) were explored to determine the surfactant composition that achieved stable nanoscale emulsions. Emulsification was conducted at 80 ± 2 °C for 10 min using an oil bath and mixing with a magnetic stirrer to form a water-in-oil emulsion. The samples were removed from the oil bath and cooled to below their PIT, with continuous mixing to room temperature (20 ± 2 °C), to obtain an oil-in-water emulsion. Ratios 3:1 and 1:0 of Kolliphor ® RH40 and Span ® 80, respectively, resulted in homogeneous emulsions ( Fig. S1).
The droplet size and polydispersity index (PDI) of emulsions obtained with these two ratios were determined (Zetasizer Nano-ZS, Malvern Instruments) by diluting with HPLC grade water (10% v/v) and readings were taken at 22 °C, with a scattering angle of 173° using polystyrene disposable cuvettes. Kolliphor ® RH40 and Span ® 80 at 3:1 ratio gave smaller and more defined particle size (96 nm, PDI 0.20) than 1:1 ratio (130 nm, PDI 0.24) and therefore was selected for emulsion preparations.

Fig. S2
TEM images (Scale bar = 500 nm) of the nanocapsules showed that the eugenol and clove oil containing nanocapsules had a core-shell structure, indicating potential encapsulation of the antimicrobials. The blank nanocapsules did not have a distinct core and shell.

Fig. S3
Particle size and polydispersity index of the nanoemulsions and nanocapsules.

Fig. S4
Flow cytometry histograms showing the population shift of nanocapsules to higher fluorescence intensity when the Dil dye is incorporated in the formulation, confirming encapsulation of the dye.

Preparation of PA13 and PA155 coated coverslips
Polymers were spin-coated onto circular glass coverslips (19 mm diameter). 75 µL of 2% w/v polymer solutions in tetrahydrofuran were spin-coated at 2000 rpm for 10 s using a desktop spin coater (6708D, Speedline technologies). Coated coverslips were dried in a convection oven at 40 ºC overnight and sterilised for 30 min using UV light prior to inoculation.

Preparation of Crosslinked-PA13 coated coverslips
Prior to coating, the surface of glass coverslips (19 mm diameter) was functionalised with 3-(trimethoxysilyl)propyl methacrylate. Coverslips were placed one at a time in 10% NaOH (aq) in a glass beaker and agitated gently for 5 h. The coverslips were washed thoroughly with HPLC grade water and dried in an oven at 115 °C. The coverslips were transferred to a high density polyethylene container containing 30 mL acetonitrile, 3 mL trimethylamine and 6 mL with 3-(trimethoxysilyl)propyl methacrylate and agitated gently overnight at room temperature. The coverslips were washed with acetone (3 × 50 mL, swirled gently and decanted) and dried at 115 °C for 1 h in a glass container.
Methyl methacrylate and N,N-dimethylacrylamide were mixed in 9:1 molar ratio and 20% (w/w) 575 Da polyethylene glycol diacrylate (PEGDA-575) was added to give a monomer-crosslinker mixture. A polymerisation mixture was prepared with the monomer-crosslinker mixture (3.5 g, 70% w/w), Irgacure 2959 (0.1 g, 2% w/w), MeOH 0.9 g, 18% w/w) and water (0.5 g, 10% w/w) (for the optimisation of the composition see section 8.1). This mixture was pipetted (8 µL spots) on to acetate sheets and the 3-(trimethoxysilyl)propyl methacrylate functionalised coverslips were S8 placed gently over each spot. The acetate sheet was transferred to a UV source (CL-1000 Ultraviolet Crosslinker-UVP) and irradiated (UV-365 nm, 8 Watt, energy 1000 mJ cm -2 ) for 90 min. The acetate sheets were dried overnight under ambient conditions. The coverslips were gently removed from the acetate sheets and immersed in excess water for 30 min and dried at 60 o C for 2 h in a fan-assisted oven. UV sterilisation (30 min) of coverslips was performed prior to inoculation.

Quantification of nanocapsules in the Eugenol-Network
The total eugenol content was determined using HPLC (as detailed in ESI, section 7.1 and 7.2) to be 3.3 ± 0.2 mg (average ± SE, n = 3) of eugenol per monolith of Eugenol-Network (3 mm height and 10 mm diameter cylinders). Since the nanocapsule dispersions contain 15% nanocapsules (with 10% poly(laurylacrylate), encapsulating 5% eugenol), 9.9 ± 0.6 mg of nanocapsules were incorporated per monolith. Therefore the Eugenol-Network contained 42 ± 3 µg of the nanocapsules per mm 3 of the coating. Fluorescence intensity of coverslips was measured using a well plate reader (Biotek, λ Ex 530/25; λ Ex 590/35, with 5×5, equally spaced points analysed within the bottom of the well) followed by washing with water (4 × 1 mL). % of reduction in fluorescence was measured by comparing the fluorescence before and after 4 washes. The formulations with 20% crosslinker showed the least reduction in fluorescence (therefore the best retention of the nanocapsules), with the retention consistent for various concentrations of water in the polymerisation mix (Fig. S6). Visual examination of the polymerisation mixtures revealed that those with 0% and 10% water showed no phase separation. In order to obtain uniform polymerisation when used for coating, the optimised polymerisation mixture contained (w/w) 18% MeOH, 10% water, 2% initiator, and 70% monomercrosslinker mixture (with 20% crosslinker). The molar ratio of MMA, DMAA and PEGDA-575 was 86.4: 9.4: 4.2, respectively. The best retention of nanocapsules was achieved with 20% of crosslinker.

Optimisation of porogen concentration for the preparation of Porous-PA13
The optimised polymerisation mixture (Section 8.1) was mixed with 0.0%, 2.0%, 4.8%, 9.1% or 16.7% w/w of 3 kDa poly(ethylene glycol) (PEG-3000). 200 µL of these solutions were pipetted into cylindrical silicone moulds (diameter 10 mm and height 3 mm) and polymerised with UV light (1 h). The polymer monoliths were then removed from the moulds, washed with acetone and acetone/water and dried (see section 5).
Each of the polymer monoliths were placed in a well of 24 well-plate and 100 µL of Dil nanocapsules was added over each monolith. After 15 minutes of incubation and visual inspection, only the formulation containing 16.7% PEG-3000 allowed penetration of the nanocapsules (with the entire monolith appearing uniformly coloured by pink Dil nanocapsules), while the Dil nanocapsules stayed on the surface of monoliths for all other PEG-3000 concentrations.

Effect of solvent composition on the migration of nanocapsules into porous-PA13 monoliths
MeOH and N-methyl-2-pyrrolidone (NMP) were investigated for their ability to induce pore formation during polymerisation to allow penetration of nanocapsules into the monoliths/coatings.
Porous-PA13 monoliths were prepared with MeOH and NMP as the solvent and their ability to permeate the nanocapsules was compared. After incubation (4 h) of the monoliths with 100 µL of Dil nanocapsules (in a 24 well-plate), only the monoliths prepared with MeOH allowed penetration of nanocapsules (the entire monolith appeared uniformly coloured on visual inspection), while the nanocapsules stayed on the surface on the NMP based monoliths. Next, the monoliths were S12 incubated 20 h in water (2 mL in a 24 well-plate with gentle shaking) and fluorescence of supernatant (200 µL) was measured in a 48 well-plate (Biotek plate reader, λ Ex 530/25; λ Em 590/35).
The monoliths prepared with NMP showed no retention of nanocapsules after washing with water (2 mL) while those prepared with MeOH retained the nanocapsules within the matrix (Fig. S7). The washed monoliths were then incubated 10 h in water (2 mL) followed by fluorescence measurement of the supernatant. Fluorescence was measured again after another washing and incubation with water (2 mL, 30 h) (Table S2).  shaking at 300 rpm, Incushake MIDI). 100 µL of the culture was added to 10 mL of fresh LB broth (in 50 mL falcon tubes) and incubated 2 h at 37 °C with shaking (300 rpm) to obtain a sub-culture of each strain. The sub-cultures were centrifuged (Megafuge 1.0, 3000 rpm for 10 min), supernatant removed, and 1 mL of PBS added. The suspension was mixed thoroughly by pipetting up and down, vortexed, transferred to a 2 mL eppendorf tube and (13000 rpm for 1 min, Sigma 1-13). Cells were washed with PBS (2 × 1 mL, vortexing and centrifugation at 13000 rpm for 1 min) and suspended in 1 mL of PBS. Optical density (OD) measurements were used to count the bacteria (assuming 1 OD = 10 9 bacteria) by measuring absorbance at 595 nm using a WPA, UV 1101 biotech photometer and polystyrene cuvettes (10×4×45 mm, Sarsted AG & Co.).

Inhibition with nanocapsules
Nanocapsules (containing 50 mg/mL of eugenol or clove oil) were diluted in PBS (25, 10, 7.5, 5 and 1 µL of nanocapsule solution per mL). 2X Mueller-Hinton broth with 2 × 10 6 CFU/mL of MRSA and K. pneumoniae was pipetted on a 96-well plate (50 µL per well) and 50 µL of the diluted nanocapsules (or 50 µL of PBS control) were added to each well, to give final concentrations of 0.625 0.25, 0.1875, 0.125, and 0.025 mg/mL of eugenol or clove oil. The well plate was sealed with an optically clear sealing film and absorbance was measured at 600 nm (Biotek Synergy HT plate reader, endpoint kinetic, every 15 min for 16 h at 35 °C, with 5 s shaking before every reading). Background absorbance at 0 min was deducted and growth curves were constructed (Fig. S8). % growth was calculated with the formula below by comparing the absorbance (after 16 h) of the nanocapsule-treated samples with PBS-treated samples.
A t and A 0 are the absorbance of the nanocapsule-treated media at 16 h and 0 h, respectively; B t and B 0 are the absorbance of the PBS-treated media at 16 h and 0 h, respectively. % growth was plotted against the concentration of eugenol or clove oil (in the total volume of the well) (Fig. S9). IC 50 values were calculated by liner interpolation of these curves (Table S3).  where one of the samples tested (i.e., 0.025 mg/mL) gave close to 50% reduction in growth.

Inhibition with Eugenol-network
% inhibition of growth of a cocktail of MRSA and K. pneumoniae was determined. The monoliths of Eugenol-network and Blank-network (n = 3) were sterilised under UV light (20 min each side), washed with water (2 × 40 mL), and placed in a 24 well-plate. MRSA and K. pneumoniae in LB broth (2 × 10 6 CFU/mL) were added (2 mL per well) (ESI, section 9) and incubated for 24 h at 37 °C. The monoliths were removed and the absorbance (600 nm) of the media measured. After subtracting the background absorbance (media without bacteria), the corrected absorbance of Eugenol-network and Blank-network were compared and % inhibition was determined to be 64 ± 3%.

Microbial viability assays
The antibacterial activity of nanocapsule dispersions was assessed using the BacTiter-Glo™ microbial viability assay (Promega). MRSA and K. pneumoniae in 2X Mueller-Hinton broth containing 2 × 10 6 CFU/mL was pipetted in a 96-well plate (50 µL per well) and 50 µL of the S16 diluted nanocapsules (50, 25 and 10 µL of nanocapsule dispersion per mL PBS) were added to each well, and the plates incubated at 35 °C for 1 h. PBS was used as a negative control and hydrogen peroxide (8% w/w) as a positive control. 50 µL from each well was transferred to an opaque 96well plate and 50 µL of the BacTiter-Glo TM reagent (prepared as per manufacturer's instructions) was added. After sealing the well plate with an optically clear sealing film, luminescence was recorded (Biotek Synergy HT plate reader, gain = autogain, after shaking 5 min). Background luminescence was subtracted (media without bacteria) and the relative luminescence units (RLU) plotted against sample number (Fig. S10). The coverslips were then imaged in the DAPI channel using EVOS FL microscope with a 60× objective. The images were processed using ImageJ using the Split channels function, followed by the Invert function, which provided a black image of the bacteria against a white background.

Quantification of bacterial attachment on polymer coated coverslips
Fig. S12 Polymer coated coverslips, untreated or treated with eugenol or clove oil nanocapsules (no crosslinker) were incubated with MRSA or K. pneumoniae for 24 h, and the bacteria fixed and stained with Hoechst 33342. The coverslips were imaged in the DAPI channel (λ ex/em 357/447 nm, 60× objective) and the images processed using ImageJ TM , providing an image of black bacteria against a white background ( Fig. 3 and Fig. S11). The Percentage of bacterial coverage was determined from the images (n = 2) using ImageJ TM (using Threshold and Measure functions). (A) PA155 showed 90% surface coverage for both species. PA155 with eugenol nanocapsules showed 10% and 5% coverage of MRSA and K. pneumoniae, respectively. (B) PA13 showed 20% coverage of MRSA and 2% of K. pneumoniae, whereas PA13 with eugenol nanocapsules showed 0.5% coverage of both species. (C) Crosslinked-PA13 showed 19% coverage of MRSA and 4% of K. pneumoniae, whereas Crosslinked-PA13 with eugenol nanocapsules showed 1% and 0% coverage. (D) Eugenol and clove oil nanocapsules produced a significant reduction in bacterial binding with eugenol nanocapsules providing a better performance.

Hemolytic activity of Eugenol-network
A suspension of human erythrocytes (obtained by centrifugal sedimentation of human whole blood from healthy donors (ethics approval from AMREC; 15-HV-013) was prepared (20% v/v in PBS). 4 mL of erythrocyte suspension was added to Porous-PA13, Blank-Network and Eugenol-Network in a 6-well plate and incubated at 37 °C (n = 3). PBS was used as negative control and Triton X-100 (4%) was used as the positive control. After 1 h incubation the samples were mixed thoroughly and S19 The well plates were sealed and centrifuged for 6 min at 2000 g. 100 µL of supernatant was removed from each well and hemolysis was determined by comparing the absorbance (540 nm, Biotek synergy HT plate reader) of the samples against the controls. All the polymer samples showed the same absorbance values as the negative control PBS, confirming no hemolysis (Table   S3).

Hemolytic activity of the nanocapsules
50 µL of erythrocyte suspension (20% v/v in PBS) and 50 µL of the nanocapsules (no crosslinker) in PBS were added to a 96-well plate, to give final concentrations of 0.625 0.25, 0.1875, 0.125, and 0.025 mg/mL of eugenol or clove oil and incubated at 37 °C (n = 3). PBS was used as negative and Triton X-100 (4%) was used as the positive control. After 1 h incubation, the samples were mixed thoroughly and 100 µL of PBS was added to each well. The well plates were sealed and centrifuged (5 min, 2000 g). 100 µL of supernatant was removed from each well and hemolysis was determined by comparing the absorbance (540 nm, Biotek synergy HT plate reader) of the samples against the controls. % hemolysis was calculated comparing the absorbance of sample against the absorbance of Triton X-100 (100 % hemolysis) and PBS (0% hemolysis). The eugenol nanocapsules showed higher hemolysis than the nanocapsules based on clove oil and the blank controls (no hemolysis) (Table S4). Table S4. Hemolysis with nanocapsules at 25-625 µg/mL of eugenol. Erythrocyte suspension (20% v/v in PBS) was incubated h with eugenol nanocapsules (no crosslinker) followed by dilution with 100 µL of PBS. After centrifugation (5 min), absorbance (540 nm) of the supernatant was measured. Percentage of hemolysis was calculated by comparing the absorbance of sample against the absorbance of controls Triton X-100 (100 % hemolysis) and PBS (0% hemolysis).