Bismuth Nanoantibiotics Display Anticandidal Activity and Disrupt the Biofilm and Cell Morphology of the Emergent Pathogenic Yeast Candida auris

Candida auris is an emergent multidrug-resistant pathogenic yeast, which forms biofilms resistant to antifungals, sanitizing procedures, and harsh environmental conditions. Antimicrobial nanomaterials represent an alternative to reduce the spread of pathogens—including yeasts—regardless of their drug-resistant profile. Here we have assessed the antimicrobial activity of easy-to-synthesize bismuth nanoparticles (BiNPs) against the emergent multidrug-resistant yeast Candida auris, under both planktonic and biofilm growing conditions. Additionally, we have examined the effect of these BiNPs on cell morphology and biofilm structure. Under planktonic conditions, BiNPs MIC values ranged from 1 to 4 µg mL−1 against multiple C. auris strains tested, including representatives of all different clades. Regarding the inhibition of biofilm formation, the calculated BiNPs IC50 values ranged from 5.1 to 113.1 µg mL−1. Scanning electron microscopy (SEM) observations indicated that BiNPs disrupted the C. auris cell morphology and the structure of the biofilms. In conclusion, BiNPs displayed strong antifungal activity against all strains of C. auris under planktonic conditions, but moderate activity against biofilm growth. BiNPs may potentially contribute to reducing the spread of C. auris strains at healthcare facilities, as sanitizers and future potential treatments. More research on the antimicrobial activity of BiNPs is warranted.


Introduction
Infectious diseases are among the first causes of death worldwide [1] and pose a burden for global health and economy, as they negatively impact major social aspects of everyday life [2]. The main threats associated with communicable diseases are the rise of drug-resistance, the limited number and diversity of available treatments, and the emergence of new pathogens [3][4][5], some of them with pandemic potential. The Candida genus is the main cause of fungal diseases [6,7]. Every year, more than 250,000 people worldwide are affected by invasive candidiasis, leading to more than 50,000 deaths, as the mortality can be as high as 40%, even when patients receive antifungal therapy [6]. Candida albicans is the main cause of Candida-related diseases; however, other Candida species also represent a health risk [8]. Among them, C. auris, an emergent pathogenic yeast, has risen as a clinical concern worldwide, as it spreads in healthcare-related facilities and has caused several outbreaks around the world [9,10].
C. auris is an ovoid-shaped yeast which has been described to be non-dimorphic, yet, some strains can display the pseudohyphae shape [11]. This yeast has been classified as an urgent threat by the Our results also show that the BiNP anticandidal activity under planktonic conditions parallels silver nanoparticles (AgNPs), which usually display MIC values within the 1 to 10 µg mL −1 range [36]. This parallelism is relevant because AgNPs are among the most potent antimicrobial nanomaterials [54]. Also, the anticandidal activity of BiNPs is comparable to that of repositionable compounds, such as miltefosine and iodoquinol (MIC = 4 µg mL −1 ), as reported previously by our group [26]. Additionally, the activity of BiNPs matches the proposed activity of the main antifungal drugs. Currently, there are not established antifungal MIC breakpoints for C. auris; however, the CDC has proposed the following MIC values to indicate resistance: >32 µg mL −1 for fluconazole and >2 µg mL −1 for both amphotericin B and caspofungin [16]. Fluconazole is one of the most used antifungal agents against Candida infections [21], but C. auris is resistant to it; however, amphotericin B and caspofungin display stronger antifungal activity against C. auris.

BiNPs Inhibit Biofilm Formation by C. auris
We evaluated the antibiofilm activity of BiNPs during the biofilm formation stage of C. auris. As biofilms confer resistance to antifungal drugs and sanitizers, it is clinically relevant to assess the ability of nanoparticles to inhibit biofilm formation. The inhibitory activity of BiNPs on the biofilm was measured using an XTT-reduction assay that measures the metabolic activity of sessile cells within the biofilms, as originally described by our group [55].
BiNPs inhibited the biofilm formation in all C. auris strains. The calculated BiNPs IC 50 values ranged from 5.1 to 113.1 µg mL −1 , with a geometric mean of 26.8 ± 24.5 µg mL −1 ( Table 2). The IC 50 values indicate that each C. auris strains had a unique susceptibility to BiNPs, as the biofilm activity Antibiotics 2020, 9, 461 4 of 15 was different in each strain. However, the strains from Clade IV were the only ones that showed a consistent behavior to the BiNP treatments. In Figure 1 dose-response curves show the antibiofilm activity against a representative C. auris strain from each one of the four clades. The individual dose-response curves for all the C. auris AR strains are shown in Figure A1 (Appendix B). It seems that the susceptibility to BiNPs among the different strains is not related to the clade. As C. auris is an emergent pathogen, the research regarding their particular physiological traits of each strain is still ongoing, particularly regarding the behavior and differences of the biofilm stages [56,57]. Moreover, the biofilm stage has remained poorly understood up to date, even in thoroughly studied Candida species such as C. albicans [58]. Antibiotics 2020, 9, x 4 of 16 activity against a representative C. auris strain from each one of the four clades. The individual doseresponse curves for all the C. auris AR strains are shown in Figure A1 (Appendix B). It seems that the susceptibility to BiNPs among the different strains is not related to the clade. As C. auris is an emergent pathogen, the research regarding their particular physiological traits of each strain is still ongoing, particularly regarding the behavior and differences of the biofilm stages [56,57]. Moreover, the biofilm stage has remained poorly understood up to date, even in thoroughly studied Candida species such as C. albicans [58].   We noted that the antibiofilm activity of BiNPs was close to the activity reported for other anticandidal agents during the biofilm formation stage. BiNPs displayed lower antibiofilm potency than that of AgNPs against C. auris biofilms [29,47]. However, the fact that BiNPs display antibiofilm activity should be considered for further research. Regarding current antifungal drugs, BiNPs display greater antibiofilm potency than fluconazole (IC 50 > 64 µg mL −1 ), but lower than caspofungin (IC 50 = 5 to 1 µg mL −1 ) and amphotericin B (IC 100 = 1 to >8 µg mL −1 ), according to Dekkerová et al., for the AR no. 0383, no. 0386, and no. 0390 C. auris strains [59].
An interesting observation is that the biofilm activity -as measured by the XTT absorbance− displayed an irregular behavior in response to the BiNP treatments ( Figure A1 (Appendix B)). Some C. auris strains exhibited an increase in the biofilm activity-higher than the untreated control-when treated with sublethal concentrations of BiNPs. After the peak of intensity-up to 2.5 times higher than in the control-the activity declined again as the BiNP concentrations increased.

BiNPs Alter the Cellular Morphology and Structure of C. auris Biofilms
To assess the effects BiNPs on the structural organization of C. auris biofilms and cell morphology, we analyzed the untreated and the BiNP-treated samples via scanning electron microscopy (SEM). All the strains of C. auris, from the four main clades, were exposed to sub-inhibitory concentrations of BiNPs during the biofilm formation phase. Then, we visualized the impact of the nanoparticles on the fungal cell morphology and biofilm structure. As the different strains exhibited variations in their susceptibility to BiNPs, the concentration of BiNPs used for the SEM analysis was selected according to their corresponding calculated IC 50 values (Table 2) and the experimental concentration from the antibiofilm assays (Section 2.2). Therefore, for each strain, the selected BiNP subinhibitory treatment was the higher experimental concentration of BiNPs closest to the calculated IC 50 values.

Clade I (South Asia Clade)
SEM images revealed that the untreated strains form clade I displayed two major structural behaviors. The untreated samples cells first subgroup (AR strains no. 0388, no. 0389, and no. 0390), exhibited both the yeast and pseudohyphae morphologies within the biofilms, whereas the cells from the second subgroup (strains no. 0382 and no. 0387) only displayed the yeast-like shape. The presence of the pseudohyphae-like phenotype has been also described in other works [17]. Subgroup 1: Untreated samples from this subgroup formed biofilms that extensively covered the surface in the bottom of the well (Figure 2A-C), whereas in the BiNP-treated samples, the covered area by the biofilm was noticeably reduced in all strains, regardless of the BiNP concentration in the different samples (from 16 to 128 µg mL −1 ) ( Figure 2D-F). The dominant shape for the no. 0390 strain is the pseudohyphae (Figure 2A), whereas, for the no. 0388 and no. 0389 strains, the yeast morphology is the most common ( Figure 2B,C). On the BiNP-treated, the presence of the pseudohyphae morphology was reduced in all strains ( Figure 2D-F). Subgroup 2: Similar to the subgroup 1, untreated samples also form biofilms that extensively cover the surface in the bottom of the well ( Figure 3A,B), whereas the BiNP-treated samples exhibited an evident reduction in all strains, regardless of the BiNPs concentration (16 and 128 µg mL −1 ) ( Figure 3C,D). In contrast to the subgroup 1, the cells from the subgroup 2 (strains no. 0382 and no. 0387), only display the yeast-like shape, with no evidence of the pseudohyphae-like shape ( Figure 3A,B). The BiNP-treated samples preserved the morphology of the cells, although some alterations on cell size and shape were observed ( Figure 3C,D).

Clade II (East Asia Clade)
The AR no. 0381 strain is the only one from this clade in the CDC panel. The biofilms from the untreated samples partly covered the bottom of the well ( Figure 4A), and the cells display the typical yeast-like morphology ( Figure 4A). BiNP-treated samples (64 µg mL −1 ) did not reveal a noticeable reduction in the biofilm formation ( Figure 4B); moreover, the cells did not show evident alterations on cell morphology ( Figure 4B) when compared with the untreated control.   The AR no. 0381 strain is the only one from this clade in the CDC panel. The biofilms from the untreated samples partly covered the bottom of the well (Figure 4A), and the cells display the typical yeast-like morphology ( Figure 4A). BiNP-treated samples (64 µ g mL −1 ) did not reveal a noticeable reduction in the biofilm formation ( Figure 4B); moreover, the cells did not show evident alterations on cell morphology ( Figure 4B) when compared with the untreated control.

Clade III (Africa Clade)
Untreated samples from the AR no. 0383 and no. 0384 strains formed biofilms that moderately covered the well surface ( Figure 5A, B). The cells from both strains exhibited the typical yeast-like shape ( Figure 5A, B). Subinhibitory concentrations of the BiNPs inhibited the biofilm formation in the no. 0383 strain, but the lower concentration used for strain no. 0384 (based on initial calculated IC50 values) did not result in significant inhibition of biofilm formation ( Figure 5C, D). BiNPs altered the cell morphology in both strains ( Figure 5C, D).

Clade III (Africa Clade)
Untreated samples from the AR no. 0383 and no. 0384 strains formed biofilms that moderately covered the well surface ( Figure 5A,B). The cells from both strains exhibited the typical yeast-like shape ( Figure 5A,B). Subinhibitory concentrations of the BiNPs inhibited the biofilm formation in the no. 0383 strain, but the lower concentration used for strain no. 0384 (based on initial calculated IC50 values) did not result in significant inhibition of biofilm formation ( Figure 5C,D). BiNPs altered the cell morphology in both strains ( Figure 5C,D).

Clade IV (South America Clade)
The untreated biofilms from the AR no. 0385 and no. 0386 strains extensively covered the surface of the well ( Figure 6A,B). Also, for both strains, cells within the biofilms displayed the typical yeast morphology ( Figure 6A,B). The BiNP-treated samples (BiNPs = 32 µg mL −1 for both strains) exhibited a visible reduction in the covered area ( Figure 6C,D); however, the cell morphology was mostly unaltered by the treatments.
We observed that the untreated samples from the different C. auris strains display variations on their phenotypes. This diversity of size and shapes has been described in other works [17,47]. In this work, we expand the current knowledge as we associate the observed specific phenotypes to particular clades.
On the BiNP-treated samples, we observed that nanoparticles altered the shape and size of the yeast cells in some strains. The cause of those alterations is unknown, as this is the first time that the effects of BiNPs are assessed in C. auris. Moreover, although the antimicrobial activity of BiNPs and thiolated bismuth complexes have been studied by different groups [60][61][62], their physicochemical interactions between BiNPs and microbial cells remains largely unknown. Bismuth nanoparticles likely exert anticandidal activity by their own, and also may be releasing antimicrobial bismuth ions and thiolated bismuth compounds, as it has been observed in other metallic nanoparticles [44,63]. Moreover, it has been suggested that metal-containing compounds are a promising alternative for developing novel substances with antibiotic properties [12], and the results from different research groups-including ours-increase the evidence for supporting that premise. Nevertheless, the mode of action of BiNPs and the physicochemical interactions within the nanoparticles-cells-biomolecules complex system remains yet to be addressed.

Clade IV (South America Clade)
The untreated biofilms from the AR no. 0385 and no. 0386 strains extensively covered the surface of the well ( Figure 6A, B). Also, for both strains, cells within the biofilms displayed the typical yeast morphology ( Figure 6A, B). The BiNP-treated samples (BiNPs = 32 µ g mL −1 for both strains) exhibited a visible reduction in the covered area ( Figure 6C, D); however, the cell morphology was mostly unaltered by the treatments. We observed that the untreated samples from the different C. auris strains display variations on their phenotypes. This diversity of size and shapes has been described in other works [17,47]. In this work, we expand the current knowledge as we associate the observed specific phenotypes to particular clades.
On the BiNP-treated samples, we observed that nanoparticles altered the shape and size of the yeast cells in some strains. The cause of those alterations is unknown, as this is the first time that the
Nanoantibiotics: We used PVP-coated bismuth nanoparticles (BiNPs) synthesized by a method previously reported by our group [51]. These PVP-BiNPs were produced via a fast, facile, and cost-effective chemical reduction process. Briefly, bismuth nitrate salts were added to a stirring glycine solution, pre-warmed at 75 • C; then pH was raised to 9 using NaOH. After, dimercaptopropanol and PVP solutions were consecutively added to the warm alkaline bismuth solution, still under constant stirring. Finally, NaBH 4 was added dropwise in two stages. The formation of the BiNPs is rapidly evident by the change of color of the suspension, from yellow to black. The suspension was kept on vigorous stirring for 10 min. This synthesis protocol can be easily replicated in non-specialized facilities, and nanoparticles can be produced in less than 1 h. The obtained BiNPs are small spheroids (average diameter of 8 nm), with a negative surface charge.
Strains: We used 10 different Candida auris strains, from the Centers for Disease Control and Prevention (CDC) Antimicrobial Resistance (AR) Isolate Bank stock [64]. The AR strains, organized by clade, are Clade I: South Asia-(AR strains no. 0382, no. 0387, no. 0388, no. 0389, and no. 0390); Clade II: East Asia-(AR strain no. 0381); Clade III: Africa-(AR strains no. 0383 and no. 0384); and Clade IV: South America-(AR strains no. 0385 and no. 0386). Freshly cultured C. auris strains were prepared as follows: a loopful of C. auris cells from frozen glycerol stocks were subcultured onto yeast extract-peptone-dextrose (YPD) (BD Difco, MD), at 30 • C for 48 h. Then, a couple of C. auris colonies were transferred into YPD broth and incubated at 35 • C, overnight, in an orbital shaker. Cells from these fresh cultures were used for the susceptibility assays described in the following sections.

Susceptibility Tests on Planktonic Cells
To assess the antifungal activity of BiNPs, the CLSI M27 protocol [53] was followed, with slight modifications. Briefly, yeast cells from the overnight cultures were washed twice in PBS and adjusted in RPMI culture media and transferred to a 96 multi-well round-bottom plate. BiNPs were prepared in a two-fold dilution series in RPMI and transferred to the multi-well plates with the yeast cells. The experimental final concentrations of yeast were 2.5 × 10 3 cells mL −1 , whereas the BiNPs serial dilutions ranged from 0.5 to 256 µg mL −1 . Plates with BiNP-treated cells and the controls (untreated cells, BiNPs with no cells, blank RPMI with no cells, and no BiNPs) were cultured for 48 h, at 180 rpm, 35 • C. The Minimal inhibitory concentration (MIC) was set as the concentration where no microbial growth was observed visually, according to the guidelines of the M27 protocol from the CLSI.

Antibiofilm Activity Assays
The inhibitory effect of BiNPs on the biofilm formation stage of C. auris was determined using a method previously reported by our group [55], with minor modifications. Briefly, the yeast cells from the overnight cultures were washed twice in PBS and the cells were adjusted in RPMI and transferred to flat-bottom 96-multiwell plates. BiNPs were prepared in a two-fold dilution series, then transferred to the multi-well plates with the yeast cells. The experimental final concentration for the biofilm assays was 1 × 10 6 cells mL −1 ; whereas the BiNP serial dilutions ranged from 0.5 to 256 µg mL −1 . The multi-well plates with the BiNP-treated yeast cells and the controls (untreated cells, BiNPs with no cells, blank RPMI with no cells, and no BiNPs) were cultured stationery, at 37 • C for 24 h, for inducing the biofilm formation phase.
To assess the antibiofilm effect of BiNPs, the XTT absorbance was used as a measure of biofilm activity [55], as follows: post-incubation, biofilms were washed twice with PBS and 100 µL of XTT/menadione solution were added to the plates with the treated and the controls. Immediately, the plates were protected from light and incubated for 2 h at 37 • C. The absorbance of XTT was measured at λ = 490 nm in a Benchmark Microplate Reader (Bio-Rad Inc). From the absorbance readings, dose-response curves were obtained to calculate the IC 50 , set as the concentration of BiNPs required for reducing the biofilm activity by 50%. The IC 50 was calculated by fitting the normalized results to the variable slope Hill equation (to assess the nonlinear dose-response relationship), in the software Prism 8 (GraphPad Software, Inc.).
To ensure the experimental reproducibility, the anticandidal activity of BiNPs was evaluated using different rounds of nanoparticle syntheses, using three independent replicates of the 96-multiwell plates, with two replicates of the treatments within each multiwell plate (n = 6, for each strain), for both the planktonic and the biofilm stages.

Ultrastructural Analysis
The effect of the BiNPs on the structure of the C. auris biofilms was evaluated in the inhibition of biofilm formation experiments via scanning electron microscopy (SEM). Briefly, biofilms were treated with subinhibitory concentrations of BiNPs, according to the corresponding calculated IC 50 values for each C. auris strain. After incubation, the biofilm samples were washed twice with PBS, then fixed with 2% glutaraldehyde for 3 h, and then stained with 1% osmium tetroxide for 30 min. Then, the samples were dehydrated using an ascending concentration series of ethanol, from 30% to 100%. Then, ethanol was completely removed, and the samples were left to dry overnight. The dehydrated samples were coated with gold in a sputter coater SC7620 (Quorum Technologies). To form a uniform and thick gold layer over the samples, the current was set at 25 milliamperes for 3 min. Finally, the samples coated with gold were observed in a TM4000Plus Scanning Electron Microscope (Hitachi Inc.), using the 500× and 2500× magnification, with a voltage of 10 KeV.

Conclusions
Overall, our results seem to indicate that bismuth nanoparticles display a strong anticandidal activity against all the multiple C. auris CDC AR strains when tested under planktonic conditions. However, the BiNPs displayed much more modest antibiofilm activity, with also more accused differences among the different strains, which were not necessarily related to their clade. Despite this lower activity, treatment with BiNPs affected the biofilm structure and, in some instances, the cell morphology of cells within biofilms. Our results seem to indicate that the broad anticandidal activity of bismuth nanoparticles may contribute to reducing the spread of the multidrug-resistant C. auris strains in the healthcare-related facilities. Finally, more studies regarding the antimicrobial properties of BiNPs, against different pathogens, will contribute to expanding their future potential applications. Figure A1. Susceptibility of the multiple C. auris strains, clustered by clade, exposed to BiNPs during the biofilm formation stage. The dose-response curves (continuous line) show the effect of the BiNPs (as XTT absorbance readings) during biofilm formation in the multiple strains of C. auris, clustered by clade. Within each graph, the individual points show the average absorbance % over the BiNPs concentration, whereas de vertical bars in each point represent the standard deviation (n = 6).