*2.1. Components*

The resin matrix was prepared using Urethane Dimethacrylate (UDMA) (DMG, Hamburg, Germany) as a base monomer and PolyPropylene Glycol Dimethacrylate (PPGDMA) (Polysciences, Inc., Warrington, FL, USA) as a diluent monomer. 4-methacryloxyethyl trimellitate anhydride (4-META) (Polysciences, Inc., Warrington, FL, USA) and Camphorquinone (CQ) (DMG, Hamburg, Germany) were added at low level. The filler consisted of two di fferent sizes of aluminosilicate glass of 7 μm and 0.7 μm (DMG, Hamburg, Germany), fumed silica (Aerosil OX 50, Evonik, Essen, Germany), monocalcium phosphate monohydrate (MCPM) (Himed, Old Bethpage, NY, USA) and polylysine (PLS) (Handary, Brussels, Belgium).

UDMA is a common base monomer, that has been used as an alternative to bisphenol A-glycidyl methacrylate (Bis-GMA) [15]. In previous studies, PPGDMA, a diluent monomer that can be used instead of triethylene glycol dimethacrylate (TEGDMA), enhanced paste stability, increased light activated polymerization and reduced associated shrinkage [11]. 4-META is an adhesion-promoting monomer [16]. Monocalcium phosphate monohydrate (MCPM) particles added to composites can promote water sorption and react to produce brushite crystals of greater volume thereby giving expansion to compensate polymerisation shrinkage. Furthermore, release of its phosphate ions can promote apatite precipitation from simulated dentinal fluid to remineralize dentine [17].

#### *2.2. Paste Preparation*

Four formulations with 1 variable (PLS) were prepared. The resin (liquid) phase consisted of UDMA (72 wt%), PPGDMA (24 wt%), 4-META (3 wt%), and CQ (1 wt%) [11]. This was prepared by mixing the components and stirring for 24 h at room temperature on a magnetic stirrer hot plate (Jeo Tech) until a clear liquid was achieved. The filler phase contained glass of 7 μm, 0.7 μm and fumed

silica in the ratio 6:3:1. 10 wt% MCPM was added to the filler as in F2 in previous work [11] but the tricalcium phosphate was removed. Additionally, the filler PLS level was varied (0, 0.5, 1 or 2 wt%) instead of being fixed at 2 wt%. Filler: resin were mixed in the weight ratio 5:1 for 45 s at 3500 rpm using a centrifugal mixer (Speedmixer, Hauschild Engineering, Hamm, Germany). This high ratio produced pastes with a consistency comparable with the commercial packable composite Filtek Z250 (3M, Bracknell, UK) used as an additional control in antibacterial studies.

#### *2.3. Composite Discs Preparation*

Disc-shaped specimens were formed by applying the composite pastes within metal circlips with internal diameter 10 mm and thickness of 1 mm and pressing them between two sheets of acetate (Figure 1). The specimens were then photopolymerized using a blue light-emitting diode curing unit with a wavelength of 450–470 nm and power of 1100–1300 mW/cm<sup>2</sup> (Demi Plus, Kerr Dental, Bioggio, Switzerland) with the tip in contact with the acetate sheet. The curing duration was 40 s on each side of each disc. This method gives 72% monomer conversion for the formulations [11]. The discs were then removed from the circlip, any excesses were trimmed with a no.11 blade, and stored at room temperature in the dark until required.

**Figure 1.** Preparation of composite disc and disc placed in well.

#### *2.4. Mass and Volume Change*

The mass and volume change of set discs versus time in deionised water (DW) were determined using a density kit and four-figure digital balance (AG204, Mettler, Toledo, OH, USA) according to the ISO 17304:2013. 1% sodium dodecyl sulphate (Sigma-Aldrich, Gillingham, UK) in DW was used as the buoyancy medium.

Discs (*n* = 3), prepared as above, were immersed in 1 mL of DW in individual sterilin tubes at 23 ◦C. At 1, 3, 6 h, 1, 2, 5, 7 days, 2, 4 and 8 weeks, the discs were removed from the solution, their surfaces blot dried using absorbent paper, weighed in air and in the buoyancy medium and then placed into new tubes with fresh 1 mL of DW. Initial mass and volume were calculated by extrapolation of early data versus square root of time to zero. Mass and volume change were then calculated as percentages of original mass as shown in detail in previous publications [10,11].

## *2.5. Polylysine Release*

High-performance liquid chromatography (HPLC) (Shimadzu corporation, Kyoto, Japan) was used to measure polylysine release from composite discs. The composite discs (*n* = 3) were prepared as above, weighed and then immersed in 1 mL deionized water (DW) in individual sterilin tubes. The discs were removed from the DW and placed in a new tube with 1 mL of fresh DW at 1, 3, 6 h, 1, 2, 5, 7 days and 3 weeks. The storage solutions were stored at 23 ◦C prior to analysis.

A normal phase column in hydrophilic interaction liquid chromatography (HILIC) mode was used. The mobile phase was 50 vol% acetonitrile in DW with 0.1 vol% of phosphoric acid (97%) added. Flow rate, run time, temperature, and UV detection wavelength were 1.0 mL/min, 43 min, 30 ◦C and 210 nm respectively. PLS solutions of 10 to 100 ppm were used to generate a calibration curve which was then employed to determine PLS concentrations in each storage solution. Cumulative PLS versus time, as a percentage of that calculated to be in the original specimen, was then determined [10].

#### *2.6. Bacterial Growth Inhibition*

A single colony of *S. mutans* UA159 was inoculated into brain heart infusion broth (BHI, Oxoid, Basingstoke, UK) and incubated statically at 37 ◦C for 16 h in air enriched with 5% CO2. The culture was diluted in BHI broth to generate an inoculum of 8 × 10<sup>5</sup> and 8 × 10<sup>6</sup> CFU/mL which was confirmed by viable counting.

On the day of each experiment the discs (*n* = 3) were placed in a custom made ultraviolet light box and irradiated for 30 min on each side to ensure sterility. Decontamination was verified by plating on agar. The discs were then placed into a 24 well plate as seen in Figure 1. 1 mL of the *S. mutans* inoculum was added to each well and the plates placed in the incubator on a shaking tray at 200 rpm at 37 ◦C in air. At 24 h, the cultures were 10-fold serially diluted and plated out on BHI agar (Oxoid, Basingstoke, UK). Bacterial colonies were counted after 3 days of incubation at 37 ◦C in 5% CO2.

#### *2.7. Surface Bacteria Observations*

To assess viability of attached bacteria on material surfaces, composite discs (*n* = 2) were immersed in an *S. mutans* suspension with concentration of 5 × 10<sup>6</sup> CFU/mL and incubated statically for 72 h in air at 37 ◦C. The discs were removed and gently immersed in fresh BHI broth to remove unattached bacteria before placing in a clean well plate for staining using LIVE/DEAD Viability kit (Thermofisher Scientific, Loughborough, UK). Confocal laser scanning microscopy (Radiance 2100, Biorad, Hercules, CA, USA) with an objective lens of 10×–20× magnification was used to visualize bacteria. Images were saved and processed using Lasersharp 2000 (Biorad, Hercules, CA, USA) and ImageJ (ImageJ Developers).

#### *2.8. Statistical Analysis*

Data analysis was undertaken for volumetric analysis, polylysine release and bacterial viable counts using Analysis of Variance (ANOVA) and SPSS (IBM, Armonk, NY, USA) with *p* < 0.05 considered significant. To perform multiple comparisons of multiple formulations, the Bonferroni adjustment was used.
