**3. Results**

#### *3.1. Mass and Volume*

Mass and volume change both increased proportional to the square root of time. Final mass change at 2 months increased linearly with polylysine content from 0.7 to 1.7 wt% (RSQ = 0.96). Final volume change, however, was not significantly affected by PLS level (*p* > 0.05) and was between 2.2 and 2.5 vol% (Figure 2).

**Figure 2.** Graph of percentage change in mass and volume of the three formulations containing polylysine after 2 months. Error bars = st. dev. (*n* = 3). \* indicates significance between results (*p* < 0.05).

#### *3.2. Polylysine Release*

Cumulative polylysine release was initially proportional to the square root of time but then levelled between 2 and 3 weeks (Figure 3). Cumulative percentage release was higher with 2% PLS at 1 and 2 days at 2.4 and 2.7% but within experimental error independent of concentration and 4% by 3 weeks. In Figure 4, the percentage release is converted to provide a calculated total concentration per disc that would have been released into 1 mL of storage solution by 6 h, 24 h and three weeks. At 3 weeks, the PLS release in grams was proportional to the level in the filler. At earlier time points, however, doubling the filler PLS content more than doubled the release in grams (see Figure 4).

**Figure 3.** Cumulative percentage release of polylysine (PLS) versus the square root (SQRT) of time. Error bars = st. dev (*n* = 3). \* indicates that formulation with 2% PLS has significantly higher percentage release at 24 and 48 h but not at early or later times.

**Figure 4.** Cumulative polylysine release after 6 h, 24 h and 3 weeks at 23 ◦C for formulations with 0.5, 1 or 2% PLS. Error bars = st. dev (*n* = 3). \* indicates 2% PLS containing formulation has significantly higher release at all time points compared with the other 2 formulations.

#### *3.3. Determination of Antibacterial Activity*

Colony forming units are shown in Figure 5, after 24 h incubation in air with different composite formulations for two different initial inoculum bacterial levels. Irrespective of initial inoculum level, CFU increased to 10<sup>9</sup> for all controls with no PLS. This was also observed with 0.5% and 1% PLS with the higher initial inoculum. Conversely, 0.5 and 1% PLS in the composite filler caused a 90 and

99% reduction in CFU with the lower initial inoculum concentration. With 2% PLS, the 24 h CFU count was down to 10<sup>7</sup> and 10<sup>2</sup> with high versus low initial inoculum levels, respectively. The 2% PLS formulations had statistically significantly less bacteria at 24 h when compared to all other composites at both inoculum concentrations.

**Figure 5.** Bacterial growth after 24 h with two different inoculum concentrations of *S. mutans*, 8 × 10<sup>5</sup> and 8 × 10<sup>6</sup> CFU/mL. Four different formulations and a commercial composite were tested. Plates with discs and inoculum were incubated in air for 24 h at 37 ◦C while being shaken. Error bars = st. dev (*n* = 3), \* indicates 2% is significantly different when compared to other formulations.

#### *3.4. Confocal Laser Scanning Microscopy*

Figure 6 shows confocal images of composite surfaces stained with live/dead stain. The bacteria do not form a biofilm due to the lack of sucrose but are seen as individual bacteria on the material surface. As the concentration of polylysine in the composite disc increases, an increase in the proportion of dead bacteria is seen.

**Figure 6.** From top left to bottom right 0, 0.5, 1, 2% formulations. Inoculum: *S. mutans* 5 × 10<sup>6</sup> CFU/mL. Plates with discs and inoculum were incubated for 72 h in air at 37 ◦C before staining. Stained with Live/Dead staining. (*n* = 2).
