*3.2. Antifungal Studies*

After successfully loading the drug compounds, we proceeded to conduct MIC90 and MFC studies to determine the optimum concentration to apply in disk diffusion and cytotoxicity assays. All MIC90 results (Table 3) obtained in this study are consistent with published MIC90 breakpoints where available in CLSI M60 and M61 standardised protocols [57,58]. This suggests that the test compounds have similar antifungal properties to amphotericin B, but on average, they are 69% less cytotoxic while maintaining similar antifungal profiles. This could prove beneficial in treating resistant and difficult to treat fungal infections, as compounds such as thymoquinone or ocimene, which appear to have similar antifungal profiles to amphotericin B, could be given without the harsh side effects that are often observed when administering amphotericin B.



In previous research by Abdel et al. (2013), they demonstrated that topical thymoquinone cream could be used safely against vaginal candidiasis in mice at concentrations of 10% (*w/v*) [59], which is twice the concentration used in this present study (Table 3).

The mean zones of inhibition for each compound at 50 mg/L, which is twice the concentration required as advised by CLSI for all fungi, are shown in Figure 7, and examples of these zones are shown in Figure 8. The disk diffusion assay for antifungal activity indicated that thymoquinone-loaded BC has comparable inhibitory effects against all four genera of fungi to amphotericin B with no significant difference (*p* > 0.05). However, ocimene and miramistin-loaded BC compared to amphotericin B against both *Candida* and *Aspergillus* species showed a significant difference (*p* < 0.05) in activity. Amphotericin B had a mean ZOI against all four genera of fungi of 22.53 ± 0.969 mm, while thymoquinone similarly had a mean ZOI of 21.425 ± 0.925 mm. However, the mean ZOI for ocimene and miramistin were 12.475 ± 1.536 mm and 11.875 ± 1.682 mm, respectively, as seen in Figure 8A–D.

**Figure 7.** Mean zone of inhibition graphs for all four antifungal agents at a concentration of 50 mg/L loaded into 8 mm disks of purified BC tested against *A. fumigatus*, *A. niger, C. albicans,* and *C. auris*. Disks of pure BC were used as controls. Each test was conducted in triplicate 10 times (*p* ≤ 0.05, *n* = 10, error bars = SD).

Quantification of the drugs entrapped throughout the bacterial cellulose matrices were achieved by comparing the difference in weight of lyophilised unloaded BC to lyophilised loaded BC (Table 4). The difference in weight is directly proportional to the available free compound within the solution at a concentration of 50 mg/L (twice the required concentration to maintain consistency between compound concentrations), along with the ability for the BC to absorb up to 99% its weight in liquid, thus absorbing up to 99% of solubilised compounds available. Table 4 shows that all compounds were successfully loaded and retained within the BC matrix up to 78.95 ± 17.5%. Similarly processed disks were subsequently used in MTT assays to confirm any cytotoxic effects they may exert on HEp-2 cells.

**Figure 8.** Representative disk diffusion assay plates showing the ZOI of bacterial cellulose disks loaded with amphotericin B, thymoquinone, ocimene, and miramistin against (**A**) *A. fumigatus*, (**B**) *A. niger*, (**C**) *C. albicans*, and (**D**) *C. auris* (*n* = 3).



The results from our study for amphotericin B and thymoquinone against both *Candida* species are similar to previous studies conducted; however, in these studies, the drugs were modified to be either liposomal or nanoparticulate, which may have impacted

the overall MIC/MFC concentrations [60–62]. However, it should be noted that in the mentioned research, only Randhawa et al. (2015) [61] conducted their studies using internationally recognised protocols; therefore, the studies conducted by Khan et al. (2018) and Cavaleiro et al. (2015) [62,63] would benefit from being repeated using appropriate standardised protocols. Additionally, the results obtained for the antifungal activity of thymoquinone against all four genera of fungi in our study are similar to results published by Khader et al. (2009) [64]. They also showed that the MIC/MFC for yeasts ranged from 1.25 to 0.08 <sup>μ</sup>g/mL and for non-dermatophyte fungi ≥10−<sup>5</sup> <sup>μ</sup>g/mL [64]. Their results may be slightly lower than our MIC/MFC data, as the study mentioned above used clinical isolates, which could show reduced resistance to the compounds.

The results show evidence that bacterial cellulose loaded with thymoquinone, ocimene, or miramistin display antifungal activity against different species of *Candida* and *Aspergillus*. This is especially poignant as both *Candida* species and *Aspergillus* species can develop resistance to commonly used antifungal drugs [5].

#### *3.3. Cytotoxicity Studies*

Once minimum fungicidal concentrations were determined, we proceeded to conduct cytotoxicity studies with concentrations of each respective drug ranging from 80 to 10 mg/L encompassing all values (Figure 9A–E) to find the most potent concentration while maintaining an acceptable level of toxicity toward HEp-2 cells. It is also shown through the cytotoxicity assays that all three test compounds have significantly lower (*p* < 0.01) cytotoxic effects against HEp-2 cells in comparison to amphotericin B; cell viability for 50 mg/L solutions of amphotericin B was 29.25 ± 1.708%, whereas for thymoquinone, ocimene, and miramistin, it was 71.25 ± 3.594%, 65.5 ± 4.435%, and 42.5 ± 8.266%, respectively.

As a result of MIC/MFC assays, it was determined that a concentration of 50 mg/L (*w/v*) would be used in disk diffusion assays owing to an average cell viability rate of Hep-2 cells of 40% to 60% and because standard operating protocols advise using twice the drug concentration of the highest MIC. It is also worth noting that Figure 9E compiles cell viability rates for all drugs at a concentration of 50 mg/L (*w/v*); a significant difference (*p* < 0.05) can be seen in the survivability of HEp-2 cells when treated with thymoquinone, ocimene, and miramistin in comparison to amphotericin B, which showed a mean cell survival rate of 29.25 ± 0.854%. In contrast, thymoquinone at a concentration of 50 mg/L (*w/v*) showed a mean survival rate of 71.25 ± 1.797%. As shown by the 8 mm BC disk compound quantification assay, we can anticipate that up to 78.95 ± 17.5% of the free compound in solution will also be absorbed by the 4 mm BC disks used for the cytotoxicity assay. These data are supported by Khader et al. (2009), who concluded that ≥20 uM concentration of thymoquinone in vitro caused 6.37 ± 0.75% necrosis in hepatocytes [64]. The in vitro toxicity of miramistin was also supported by Osmanov et al. (2020) [65], who concluded that there were no cytotoxic effects seen at a concentration of 1000 mg/L against McCoy mammalian cell lines. Subsequent confocal microscopy of the HEp-2 cells after being exposed to the antifungal agents for 24 h can be seen in Figure 10A–E. The typical morphology (triangular) of HEp-2 cells can be observed in Figure 10E (control) with all previous images (Figure 10A–D) showing treated cells.

Figure 10A shows cells treated with amphotericin B, and as expected, they became detached from the base of the well and died, as evidenced by the circular appearance rather than being triangular in nature. Figure 9B–D show HEp-2 cells treated with thymoquinone, miramistin, and ocimene, respectively, and both show a positive correlation to the survivability data in Figure 9A–E. The vast majority of cells have retained their triangular appearance and have remained attached to the well's base, suggesting cellular survival, which is in accordance with Uribe et al. (2013), who described the cytotoxic effects of amphotericin B in myofibroblast cell lines [66].

**Figure 9.** MTT cytotoxicity assay with varying concentrations of each respective drug, (**A**) amphotericin B, (**B**) miramistin, (**C**) ocimene, and (**D**) thymoquinone, against HEp-2 cells, (**E**) condenses data of 50 mg/L drug concentration from each graph (**A**–**D**) for ease of comparison. All tests were carried out in triplicate. Control was DMSO and RPMI-1640 (50:50) (*p* ≤ 0.01, *n* = 10, error bars = SD).

Additionally, it would be interesting to investigate the antifungal agents in vivo against animal models, which would elucidate the antifungal agents' real potential as the hemocompatibility, along with biocompatibility of the overall biofunctionalised material system, could be collated. Moreover, future studies could be performed using clinical isolates to reduce the risk of selection bias. Secondly, further investigation into the mechanism of action of miramistin, ocimene, and thymoquinone would allow for a greater understanding of how these drugs exert their antifungal properties. Researchers have reported that thymoquinone and ocimene potentiate the mode of action of antibiotic compounds; however, research toward ocimene in this area is still lacking [67–69]. Nevertheless, as the studies conducted by Goyal et al. (2017), Liu et al. (2019), and Sarah et al. (2019) [67–69] were outside the remit of internationally recognised protocols, their results should be reconfirmed utilising appropriate standard protocols such as those produced by CLSI. Still,

the test compound results for the antifungal potential in our study show promise and merit further investigation.
