Three C. albicans strains were studied: a clinical isolate (CI) collected from a patient at the Santa Maria Hospital (Lisbon, Portugal), SC5314/ATCC MYA-2876 (WT) and SC5314 CAI4 ura3Δ::imm434/ura3Δ::imm434 (Δgcs), congenic to SC5314, kindly provided by Dr. Dirk Warnecke (Institut fur Allgemeine Botanik, University of Hamburg, Germany). On Δgcs, the glucosylceramide synthase gene (HSX11) was disrupted, making the strain deficient on glucosylceramide lipid (Leipelt et al., 2001). Strains stocks were kept at −80°C, with 15% glycerol. Cells in stock, previously thawed, were inoculated onto Yeast Peptone Dextrose (YPD, Sigma Aldrich, USA) agar plates, and incubated for 48 h at 37°C. After this period, an isolated fungal colony was cultured overnight at 25°C with shaking at 180 rpm in YPD broth. During this period, the culture reaches the stationary phase of growth. Cells were harvested by centrifugation at 1,880 g for 10 min at 4°C, the supernatant was removed and cells were washed three times with 10 mM HEPES buffer pH 7.4 with 150 mM NaCl, for planktonic studies, and with 10 mM phosphate buffered saline (PBS, 2.7 mM potassium chloride, 137 mM sodium chloride) pH 7.4 for biofilm assays. Afterwards, cell concentration was determined and the initial suspension was diluted to the concentration necessary for each experiment.

In vitro antifungal susceptibility tests were performed to determine the minimal inhibitory concentration (MIC). It was determined according to recommendation of the National Committee for Clinical Laboratory Standards (National Committee for Clinical Laboratory Standard, 1997), by the microdilution method, in 96-well microplates (Brito et al., 2010; Eksi et al., 2013). RPMI 1640 medium with L-glutamine (Gibco-Life Technologies, UK) was used, supplemented with 0.2% glucose, 165 mM MOPS (3-morpholinopropanesulfonic acid) (AppliChem, Germany) and buffered to pH 7.4. For the sake of comparison and as positive control, the conventional antifungal drugs amphotericin B (AMPH B) and fluconazole (FCZ) (Sigma Aldrich, USA) were also tested against the selected strains. Each well was prepared to a total volume of 200 μl, with the growth medium (RPMI), different AMPH B or FCZ concentrations (0.001 to 100 μg/ml) and C. albicans (2 × 10³ cells/ml). Controls without antifungal were also tested. Plates were incubated for 48 h at 37°C, after which optical density was measured at 540 nm. MIC was defined as the minimal concentration of a drug that, after incubation, causes 100% growth inhibition of an organism (Andrews, 2001). Experiments were performed in triplicate and values were analyzed with GraphPad Prism 5, using the Gompertz equation for MIC determination (Lambert and Pearson, 2000). Experiments with Psd1 were performed at 90 μg/ml (20 μM) (de Medeiros et al., 2014).

The formation and susceptibilities of C. albicans biofilms were determined by 96 well-plate based method (Pierce et al., 2008). To determine the optimal cell concentration for biofilm formation for each strain, 100 μl C. albicans suspension cells in RPMI 1640 with glucose 2% and 165 mM MOPS, at 1.0 × 10⁵, 1.0 × 10⁶, 1.0 × 10⁷, and 1.0 × 10⁸ cells/ml were placed on a polystyrene 96 well plate, each sample in triplicate. Plates were incubated for 12, 24, 48, 72, and 96 h at 37°C, in order to establish the optimal cell concentration for biofilm development. At the end of each step-time (incubation time), the biofilm was washed three times with PBS to remove planktonic and/or no adherent cells. At this point, 100 μl of XTT/menadione solution (1 μl menadione 1 mM in 10 ml XTT 0.5 g/l) were added on each well-plate (where the biofilm is formed) and incubated for 2 h at 37°C. After this time, an orange color reveals the metabolic activity of the cells within the biofilm. The supernatants were transferred to a new plate and the optical density measured at 490 nm.

Planktonic cells imaging were performed for all C. albicans strains as follows, 1 × 10⁵ cells/ml were incubated at 25°C in HEPES buffer, with agitation, for 6 and 24 h. AMPH B, FCZ, and Psd1 final concentrations were equal to the MIC and 10-fold higher than the MIC. As a control, cells without antifungal treatment were used. A 100 μl droplet of each test sample was applied onto a poly-L-lysine (PLL)-coated coverslip and left at 25°C for 2 h. After deposition, the samples were rinsed 10 times with filtered (0.2 μm) deionized water and air-dried at 25°C.

Untreated and treated cells were imaged using a JPK NanoWizard II atomic force microscope (JPK Instruments, Berlin, Germany) mounted on a Zeiss Axiovert 200 inverted microscope (Carl Zeiss MicroImaging, Jena, Germany). Measurements were carried out in intermittent contact mode, at room temperature, using uncoated silicon ACL cantilevers (Applied NanoStructures, Mountain View, CA, USA). These cantilevers have typical resonance frequencies of 145–230 kHz and spring constants of 20-90 N/m. The scan rate was set to less than 1 Hz for imaging and image resolution was set to 512 × 512 pixel for all images. Height, error signal and phase contrast images were recorded, and line-fitted as required. From recorded images, height and size information was obtained with the JPK Data Processing software v.4.2.53.

Roughness analysis of AFM height images was performed using the Gwyddion 2.31 software (Czech Metrology Institute, Brno, Czech Republic). Roughness was calculated from the root mean square value (RMS, i.e., standard deviation of the distribution of heights over a 1 × 1 μm² imaged area). The results of this processing were statistically analyzed using analysis of variances (ANOVA) and Bonferroni post-tests.

C. albicans washed suspensions were incubated at room temperature, with agitation, for 24 h. AMPH B, FCZ, and Psd1 final concentrations were 10-fold higher than the MIC. Final cell concentration was 1 × 10⁵ cells/ml and as control, cell samples were incubated without any treatment. 100 μl of each sample were placed onto a poly-L-lysine (PLL)-coated glass coverslip and left at 25°C for 2 h. After deposition, samples were rinsed 10 times with HEPES buffer to remove any cells that had not adhered to the coverslip. 100 μl of HEPES buffer were added to the adhered cells to avoid sample drying. Measurements were carried out in HEPES buffer at 25°C, using 200 μm long gold reflex coated silicon-nitride OMCL-TR400PSA-1 cantilevers (Olympus, Japan). These cantilevers have typical resonance frequencies of 8–14 kHz and spring constants of approximately 0.02 N/m.

First, to have a prior overview of the cells shape and height, force maps were performed using a 10 μm/s approach and retraction speed, Z length of 3 μm and a relative set-point of 0.4 V. The coordinates on the map were then chosen. Afterwards, one location per each cell was chosen and force-distance measurements were conducted over those coordinates, in triplicate, using a 3 μm/s approach and retraction speed, Z length of 3 μm and a relative set-point of 0.4 V. These conditions ensure that the identation ranged from 5 to 10% of cells height. Retraction force-distance curves were processed with the JPK Data Processing software v. 4.2.53. After processing, the four-sided pyramid Hertz modified equation was applied to the curves and the Young's modulus obtained. The results of this processing were statistically analyzed using ANOVA and Bonferroni post-tests.

Biofilm inhibition and eradication assays were determined by using the cells conditions obtained in biofilm development assays. Once the cells concentrations for each strain were determined, concentrations of AMPH B, FCZ, and Psd1 used for biofilm inhibition and eradication assays correspond to 10 and 100-fold higher than the MIC. In inhibition assays, the antifungal drugs were placed at the same time than cells (pre-mixing antifungal with cells) and incubated during the same step-time previously determined for biofilm development. For eradication assays, antifungals were added once the biofilm was formed, for each strain, and left to incubate for 24 h at 37°C. In both cases, biofilms medium were replaced with PBS for microscopy measurements. For AFM imaging, biofilms were washed 10 times with filtered deionized water (0.2 μm) and air-dried at room temperature. Biofilm images for formation, inhibition and eradication of C. albicans strains were measured directly on the surface where cells were grown.

Live/Dead assay kit was used both on C. albicans biofilms and plancktonic cells. LIVE/DEAD® FungaLight™ Yeast Viability Kit (L34952, LifeTechnologies, USA) is composed by two fluorescent probes, SYTO 9 and propidium iodide (PI). In a population of live and dead cells, SYTO 9 nucleic acid labels all yeast in a population, those with intact membranes and those with damaged membranes. In contrast, PI penetrates only yeast with damaged membranes, causing a reduction in the SYTO 9 stain fluorescence by fluorescence resonance energy transfer (FRET) when both dyes are present (Johnson and Spence, 2010). As a result, yeast with intact membranes is stained with green fluorescence, whereas yeast with damaged membranes is stained with red fluorescence.

Optical microscopy experiments with a Zeiss LSM 510 META confocal point-scanning microscope (Jena, Germany) were carried out in order to examine the architecture and the viability of the cells before and after exposure to antifungal agents. Argon (488 nm; 45 mW) and diode-pumped solid-state (561 nm; 15 mW) lasers were used with a 40 × dry-objective. Cells were incubated with AMPH B (10 mg/ml), FCZ (40 mg/ml), or Psd1 (900 μg/ml) for 24 h prior to imaging. These antifungal concentrations were selected to be 10 times more than the planktonic MIC. Afterwards, biofilms were labeled with SYTO 9 and PI probes, incubated for 15 min and images acquired. Images were analyzed with ImageJ 1.47v (rsbweb.nih.gov/ij/).

For flow cytometry assays, C. albicans strains were incubated for 24 h with AMPH B, FCZ, and Psd1 with antifungal concentration equal to the planktonic MIC and 10-folder higher. The double labeled cells were considered as positive result for death, since the green dye is present in all cells, and besides the FRET phenomena, the green fluorescence intensity had always a small contribution. C. albicans stains were performed according to the manufacturer's instructions. Briefly, 1 × 10⁵ cells/ml in HEPES buffer were stained with both dyes to a final concentration of 3.34 and 20 μM (SYTO 9 and PI, respectively). All samples were kept at room temperature in the dark for 15 min before flow cytometry analysis. Experiments were performed in a BD Accuri C6 Flow Cytometer (BD Biosciences, San Jose, CA, USA), using blue (488 nm) and red (640 nm) lasers to excite stained cells. Green fluorescence emission was detected with a 530 nm bandpass filter and red fluorescence emission was detected with a 670 nm bandpass filter. Fluorescence emission was acquired in bioexponential scale, and data were collected for 40 000 cells. All flow cytometer results were analyzed using FlowJo Software version 10.0x (Tree Star Inc., Ashland, OR, USA).

The activity of AMPH B and FCZ against the three C. albicans strains was determined by measuring their susceptibility to the antifungal drugs, as shown in Figure S1 (Supplementary Material). The obtained curves were fitted and MIC values were obtained using the Gompertz equation (Lambert and Pearson, 2000). The values used for Psd1 were reported in previous works (20 μM) (de Medeiros et al., 2014). As shown in Figure S1, MIC values for AMPH B and FCZ differed depending on the strain studied. Altogether, the three C. albicans strains were sensitive to AMPH B, FCZ, and Psd1. AMPH B and FCZ showed similar MIC in WT and Δgcs planktonic cells (these four MICs are all on the 3.7–4.0 μg/ml range). Lower values were obtained when CI was tested.

Biofilm formation is dependent of cellular adherence to the growth surface. To study the mode of action of Psd1 on C. albicans biofilm formation, polystyrene surfaces were selected. To optimize the conditions for biofilm formation, XTT assays were performed for the three C. albicans strains studied. The time and the cell concentrations that ensure the best biofilm growth are shown in Figure S2 (Supplementary Material). As it can be seen, the cell density needed for biofilm development differs from strain to strain. WT and Δgcs need a cell concentration 10-fold higher than the CI (10⁵ cells/ml) to initiate biofilm development. Despite this, the clinical isolate and WT formed biofilm after 24 h, while the mutant strain needs 72 h to ensure biofilm development and adherence.

AFM imaging on intermittent contact mode was used to evaluate the effects suffered by C. albicans planktonic cells after 6 h and 24 h of incubation with AMPH B, FCZ or Psd1 at the MIC and at 10-fold higher concentrations. The error signal is the difference between the detector signal and the setpoint, and provides images with greater spatial detail. In general, antimicrobial treatments induced morphological changes in all cells. The severity of these effects increased in a time-dependent manner. After 6 h, it was possible to observe small irregularities in the cell surface (blebs) and small vesicles deposited over and around the cells at 10-fold higher than the MIC antifungal concentration (data not shown). As seen in Figure 1 for CI and WT strains, untreated cells (controls) have a smooth surface and regular shape (Figures 1A,D). In contrast with CI and WT, untreated Δgcs cells showed irregular surface (Figure 1G). The incubation for 24 h with Psd1 at MIC concentration showed some deformations at cell surface (Figures 1B,E,H). At 10 × MIC enhanced effects than those obtained at MIC were observed (Figures 1C,F,I). Small blebs are being released from the cells accompanied and internal content released (Figure 1 and Figure S7).

Unlike AMPH B and FCZ (Figures S5, S6), Psd1 effects appear to be more severe. The consensual outcome is the release of the cell internal content or cells completely covered by blebs. For WT treated cells, the height of a bleb is approximately 14 nm (Figure S3). Blebs accumulate over each other, forming a bulk structure on top of the cell. Similar results were obtained for mutant cells (Figure S4). Only for CI cells a peel-off-like morphology was observed at Psd1 10 × MIC (Figure 1C). Of the three treatments, AMPH B is the one that induced a more extensive cell deformation: cells lose volume and membranes appear rougher. The clinical isolate seems to be less affected by FCZ, at both times of incubation and concentration of drug used, only with some irregularities in the cell surface appearing after 24 h of incubation.

Surface roughness was evaluated for the three C. albicans strains before and after treatment with AMPH B, FCZ, and Psd1 (Figure 2), both after 6 h and 24 h of cell treatment. There was an overall increase of the average roughness upon any of the treatments performed with the planktonic cells. These results, obtained with the RMS formula applied to 1 × 1 μm² images of the surface of the cells, are in agreement with the previous qualitative observations that surface roughness increases after antifungal treatment (Figure 1). Regarding the control conditions, CI and WT cells have lower surface roughness values (Figures 2A,D, control), below 5 nm, whereas Δgcs cells present surface roughness values above 5 nm (Figure 2G, control). WT cells were the most affected (Figures 2D–F) and Δgcs the least affected (Figures 2G–I) by any of the treatments. All treatments with AMPH B caused a statistically significant increase of membrane roughness on CI and WT cells (Figures 2A,D). This was not always the case for Δgcs cells, as their increase in roughness was not statistically significant (Figures 2H,I). When comparing the three cells treated with FCZ (Figures 2B,E,H), less effects on surface roughness were observed for CI and Δgcs. On the contrary, WT cells were strongly affected by this treatment and all conditions tested resulted in a statistically significant increase in surface roughness. Finally, Psd1 increased CI cells roughness in a way similar to AMPH B (Figure 2C), whereas for WT cells the effects of Psd1 (Figure 2F) were similar in magnitude to those of FCZ. Again, as it was observed for AMPH B and FCZ, Δgcs were the least affected by Psd1; yet, the defensin was able to strongly increase the surface roughness of these cells after 24 h of incubation, with a peptide concentration 10-fold higher than the MIC (Figure 2I).

Changes in the stiffness of the cells were assessed in two different ways. One was based on the determination of the Young's modulus of the membrane, using AFM-based force spectroscopy (for the three treatments; Figure 3); and the other was based in the observation of AFM phase-contrast images of the cell surface, which allow to visualize and distinguish softer and stiffer areas in the membrane (for Psd1 only; Figure S7).

Young's modulus determination was performed after 24 h of incubation with an AMPH B, FCZ or Psd1 concentration 10-fold higher than the MIC (Figure 3). CI cells had a mean value of membrane stiffness of 354 ± 14 kPa, WT had a mean value of 384 ± 14 kPa and Δgcs had a mean value of 315 ± 21 kPa. When Δgcs cells stiffness was compared with WT, there was an 18% reduction (red asterisk in Figure 3; ^*p < 0.05). The percentages of stiffness reduction relative to the control were calculated to better understand the different impacts of each treatment (values in the dark red box on the bottom of Figure 3). AMPH B effects on cell stiffness were more severe for Δgcs cells, with a 60% reduction of the cell initial stiffness (Figure 3). WT cells were the less affected by this treatment, with a 23% reduction of the initial stiffness. FCZ was the treatment with the least effect on the three strains and its highest reduction on stiffness was registered for the WT cells, where FCZ caused a reduction to nearly half of the initial stiffness. From the three treatments, Psd1 caused the largest reduction on CI (67%) and WT (57%) cells stiffness, whereas for Δgcs cells this was the treatment with a lower effect (34% stiffness reduction). In all cases, the treatment with the antifungal drugs, including Psd1, lead to statistically significant decreases on the cells stiffness (Figure 3; p < 0.001).

By phase contrast imaging of C. albicans after Psd1 treatment, it was possible to distinguish some changes related to sample properties such as stiffness and softness (Magonov et al., 1997; Martinez and Garcia, 2006; Garcia et al., 2007; Nie et al., 2011). Phase contrast images of control cells of the three strains all present a homogenous surface. For the three strains, the results observed are roughly the same. Blebs in the cell surface caused by the treatment with the peptide are softer than the surrounding cell surface (Figure S7, phase contrast images in Supplementary Material), and even when there are no blebs formed, these images allow to distinguish softer and stiffer areas.

Flow cytometry experiments were carried out in order to determine if Psd1 kills planktonic C. albicans cells, by live/dead staining. CI, WT, and Δgcs cells were incubated for 24 h with a concentration of AMPH B, FCZ or Psd1 equal to the MIC and 10-folder higher. The quadrants (unstained, live, dead and double positive cells) were established for each strain using their respective controls (data not shown).

Figure 4 shows the flow cytometry dot plots obtained for CI, WT, and Δgcs planktonic cells after treatment with Psd1 at MIC and 10-folder higher. Here, Psd1 had less of an effect on CI cells, with fewer cells dead in the presence of the peptide (Figures 4A–C). For WT and Δgcs, the percentage of dead cells increased to 70.6% and 23.8% in presence of Psd1 at MIC concentration, respectively. Curiously, when peptide concentration was increased to 10-folder higher, the percentage of dead cells seems to be similar for Δgcs cells (27.8%), but not for WT cells (48.7%).

Comparing all these results with those obtained for AMPH B and FCZ (Figures S8–S10), it is possible to infer that Psd1 had a stronger effect in WT and Δgcs cells (Figures 2E,F,H,I, respectively) similar to AMPH B (Figures S9, S10). AMPHB and FCZ antifungals had the same effect in CI cells: approximately 20% of cells were killed by their effect (Figures 4B,C, and Figure S8). Psd1 and AMPH B had the same behavior in all three strains: less cell death was induced in CI strain by antifungal action, while increasing death was observed in WT and Δgcs cells. On the contrary, FCZ showed less dead cells (less than 30%) in C. albicans strains.

AFM was used for imaging the effect of AMPH B, FCZ, and Psd1 on C. albicans biofilms after incubation for 24 h at a concentration 10 times higher than the MIC for planktonic cells (Figure 5). Inhibition of biofilm development was observed for CI and WT strains when treated with AMPH B, FCZ or Psd1 (data not shown). The inhibition of biofilm development showed absence of hyphae or pseudohyphae. Budding yeast cells in small groups (4–8 cells) were observed for all antifungal treatments. In the presence of AMPH B, CI seems to be less affected when compared to WT. Images of the CI biofilm (Figure 5B) show an absence of both pseudohyphae and hyphae. The effect of AMPH B was more evident on WT biofilm (Figure 5F): large deformed cells together with cells of reduced size. Both for the CI and WT, the presence of surface uncovered by the cells evidences changes in the biofilm biomass upon AMPH B action. These effects were more remarkable with FCZ and Psd1. The uncovered surface area was increased and the loss of volume was observed for CI and WT strains. FCZ at 10 × MIC induced a decrease on cell volume and size (Figures 5C,G). Psd1 caused important morphological changes on the cell surface. For the CI treated with Psd1 at 10 × MIC, the appearing of blebs and the loss of cell volume were observed (Figure 5D). WT seemed to be less affected by Psd1. Nevertheless, the appearing of small blebs can be noticed (Figure 5H). The percentages of live and dead cells were experimentally quantified (Table S1). As shown, AMPH B is more effective on biofilm eradication, when compared to FCZ and Psd1.

To evaluate the viability of yeast cells before and after antifungal treatments, confocal laser scanning microscopy images were acquire in three biofilm “phases”: formation, inhibition and eradication of the formed biofilm. As shown in Figure 6, the biofilm formed for all strains presented a 3D architecture, consisting of a network of hyphae and budding yeast cells connected at several points. Upon quantification, it was shown that the density of dead cells once the biofilm was formed was low (approximately 10%) for all strains (Figures 6A,E,I). For the inhibition assays, no cells growth was observed after antifungal treatment. Our data indicate that AMPH B, FCZ, and Psd1 inhibited biofilm formation by reducing the rate of its development.

For eradication assays, once the biofilm was developed for each strain the antifungal was added at concentration of 10 × MIC, and incubated for 24h. The results obtained show that the biofilm formed by Δgcs strain was completed eradicated (Figure 6), contrary to CI (Figures 6B–D) and WT (Figures 6F–H), where the architecture of the biofilm suffered perceptible/noticeable changes and the amount of dead cells was increased, especially for CI treated with AMPH B (Figure 6B) and FCZ (Figure 6C), and for WT treated with AMPH B (Figure 6F).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.