Evaluation of Biofilm Formation in Candida tropicalis Using a Silicone-Based Platform with Synthetic Urine Medium

Molecular mechanisms of biofilm formation in Candida tropicalis and current methods for biofilm analyses in this fungal pathogen are limited. (2) Methods: Biofilm biomass and crystal violet staining of the wild-type and each gene mutant strain of C. tropicalis were evaluated on silicone under synthetic urine culture conditions. (3) Results: Seven media were tested to compare the effects on biofilm growth with or without silicone. Results showed that biofilm cells of C. tropicalis were unable to form firm biofilms on the bottom of 12-well polystyrene plates. However, on a silicone-based platform, Roswell Park Memorial Institute 1640 (RPMI 1640), yeast nitrogen base (YNB) + 1% glucose, and synthetic urine media were able to induce strong biofilm growth. In particular, replacement of Spider medium with synthetic urine in the adherence step and the developmental stage is necessary to gain remarkably increased biofilms. Interestingly, unlike Candida albicans, the C. tropicalis ROB1 deletion strain but not the other five biofilm-associated mutants did not cause a significant reduction in biofilm formation, suggesting that the biofilm regulatory circuits of the two species are divergent. (4) Conclusions: This system for C. tropicalis biofilm analyses will become a useful tool to unveil the biofilm regulatory network in C. tropicalis.

Biofilms are mixed communities of microbes that adhere to a surface and are embedded within an extracellular matrix comprised of polysaccharides, proteins, and extracellular DNA [12]. Biofilms can grow on in-dwelling medical devices and therefore are highly associated with virulence and drug resistance, thereby hampering clinical treatment [13,14]. The formation of mature biofilms in C. albicans requires four major distinct steps: early, intermediate, maturation, and dispersal [12]. During the early stage, planktonic cells are present as in the yeast phase and must adhere to a surface. Later, pseudohyphae and hyphae are formed. The increased hyphal growth further promotes extracellular matrix production, leading to biofilm maturation and eventual dispersal [15]. Central to the understanding of the mechanisms underlying C. albicans biofilm formation is the regulation by a complex regulatory network in which several major transcriptional factors are involved in biofilm growth [16][17][18]. Here, we focus on six major regulators, Bcr1, Brg1, Efg1, Rob1, Ndt80, and Tec1 [16]. Loss of any one of these regulators in C. albicans significantly compromises biofilm formation in vitro and in vivo [16].
In C. tropicalis, a flow model has been used to examine C. tropicalis biofilm formation on latex and silicone catheters using synthetic urine (SU) medium [19]. Bovine serum-and Spider-induced biomass dry weight to determine C. tropicalis biofilms on the bottom of polystyrene plates have also been reported [20]. The former assay requires additional equipment; the latter analysis highly depends on the wash step that may cause variability with different scientists. Hence, to obtain optimal biofilm formation with C. tropicalis, several culturing conditions for biofilm development were evaluated in this study. We found that using SU in the adherence step enables C. tropicalis to form biofilms on silicone. Moreover, we showed that presence of magnesium in SU is necessary for proliferation and biofilm formation by C. tropicalis. Furthermore, deletion of the BCR1, BRG1, TEC1, EFG1, or NDT80 genes but not the ROB1 gene in C. tropicalis reduced biofilm biomass. Reintroduction of the C. tropicalis ROB1 into C. albicans rob1∆ mutant strains could not restore biofilm growth. These results suggest divergent functions of C. tropicalis ROB1 (CtROB1) compared to that of its relative species C. albicans. Overall, we established a method for C. tropicalis biofilm analysis on silicone using SU medium. This method will help the field study the molecular mechanisms of biofilm formation in C. tropicalis.

Biofilm Assay and Biofilm Staining
To develop the biofilms on the bottom of polystyrene plates, overnight YPD cultures grown at 30 • were used to inoculate polystyrene plates pretreated with bovine serum at an OD 600 of 0.5. Cells were left to adhere to the plates in Spider at 37 • and shaking at 100 rpm for 2 h. Plates were then washed with Phosphate-buffered saline (PBS) three times by pipetting after removed supernatants. Fresh Spider medium was added to each well, and plates were further incubated for 48 h at 37 • at 100 rpm [20]. Supernatants were removed, and samples were then photographed. The adhered cells were scratched out of the plastic well and quantified by measuring the OD 600 . Determination of slime index (SI) was based on the previous report with slight modifications [29]. SI was calculated from the biofilm formation values obtained by OD method in relation to the growth culture values measured by OD 600 prior to washing the wells. SI establishes a relation between biofilm formed and culture growth (SI = (biofilm/growth culture) × 100) [29].
Established protocols to measure the dry weight of biofilms in a silicone model of biofilms were performed, with slight modifications [21]. Preweighed 1.5 cm × 1.5 cm sterile silicone squares (Bentec Medical, PR72034-06N, Woodland, CA, USA) were preincubated in bovine serum (Gibco, 1861237; ThermoFisher Scientific Inc., Waltham, MA, USA) for 12 h at 37 • C and 100 rpm in a polystyrene 12-well plastic plate. Serum-treated silicone was washed with 2 mL of PBS. The washed silicone was placed in 2 mL of Spider, Lee's glucose, RPMI, yeast nitrogen base (YNB) + 1% glucose, SD, SCD + 50% serum, or SU medium for adhesion. Notably, mannitol in Spider medium was replaced with 1% glucose for C. tropicalis biofilm analyses. C. tropicalis and C. albicans cells were grown overnight in YPD medium, and approximately 1 × 10 7 cells were added on top of each silicone square. The inoculated plate was incubated with gentle agitation (100 rpm) at 37 • C for 4 h for adhesion. Silicone squares were washed with 2 mL of PBS, and incubation continued in 2 mL of fresh Spider medium for 24 h at 37 • C with gentle shaking (100 rpm). Supernatants were removed, and silicone squares were allowed to dry overnight before weighing to determine biofilm mass. Four to five replicate biofilms grown in separate wells were used for each strain.
Crystal violet staining was performed by decanting the silicone, with slight modifications [30]. Cells were stained for 30 min with 2 mL of 0.2% aqueous crystal violet per silicone. Silicone was washed with 2 mL of PBS and destained with 1 mL of 95% ethanol for 30 min. Two hundred microliters of the destain solution was placed in a 96-well plate, and absorbance at 520 nm was read using an ELISA reader (Waltham; Thermo Fisher Scientific Inc., Waltham, MA, USA). The experiment was performed in three experimental replicates.

Statistical Analyses
All statistical analyses were performed with Excel software. Differences were analyzed using a one-tailed or two-tailed Student's t-test with a 95% confidence interval.

Effects of Different Culture Conditions on Biofilm Formation in the C. tropicalis Wild-Type Strain (MYA3404)
We compared the effects of different media during the adhesion step on biofilm growth. These media included Spider, Lee's glucose, RPMI, YNB + 1% glucose, SD, SCD + 50% serum, and SU media [19,[21][22][23][24][25]. We first tested the biofilm cells of C. tropicalis on the bottom of 12-well polystyrene plates. However, the biofilm growth of C. tropicalis in different culture media was easily washed out, although cells incubated with SCD supplemented with 50% serum for adhesion showed some biofilm formation around the well and a minor amount on the bottom (Figure 1a). We further calculated the slime index (SI), where growth constitutes a correction factor in the determination of biofilm formation. Figure 1b showed that C. tropicalis exhibited significant increase in the slime index in SU medium compared to that of the Spider medium. Nevertheless, the final outcome of low quantifiable biofilms by measuring cell density in Figure 1a still caused difficulty in determination. We therefore evaluated whether these culture conditions for the adhesion stage could profoundly affect C. tropicalis biofilm formation on silicone-based materials. As shown in Figure 2, the use of RPMI 1640, YNB + 1% glucose, or synthetic urine media as the adherence medium induced the most biofilm growth, particularly with the SU medium. Thus, SU medium was primarily used to analyze C. tropicalis biofilm formation on silicone in this work.

Replacement of Spider Medium with Synthetic Urine in the Adherence Step Profoundly Affected C. tropicalis Biofilm Growth but not that of C. albicans
Previous studies have shown that C. albicans biofilms were formed in vitro using SU medium [24], and Spider medium has frequently been used for the adhesion step in biofilm development by C. albicans [16,21,22]. We therefore evaluated whether changes in the adhesion condition with SU medium affected biofilm development in both C. albicans and C. tropicalis. The results showed that both Spider and SU media were able to induce biofilm development (~12.1 and ~11.3 mg, respectively) in C. albicans (Figure 3a), whereas biomass weights with SU medium significantly increased (~7.7 mg) compared to that with Spider medium (~2.3 mg) in C. tropicalis (Figure 3a). Consistent with a previous report, EFG1 is required for biofilm growth in C. tropicalis (Figure 3a), and a mutant was used as a negative control [20]. Biofilm staining using crystal violet also showed similar conclusions, in which more biofilms were stained in SU medium than in Spider medium in the C. tropicalis wild-type strain (Figure 3b).

Replacement of Spider Medium with Synthetic Urine in the Adherence
Step Profoundly Affected C. tropicalis Biofilm Growth but not that of C. albicans Previous studies have shown that C. albicans biofilms were formed in vitro using SU medium [24], and Spider medium has frequently been used for the adhesion step in biofilm development by C. albicans [16,21,22]. We therefore evaluated whether changes in the adhesion condition with SU medium affected biofilm development in both C. albicans and C. tropicalis. The results showed that both Spider and SU media were able to induce biofilm development (~12.1 and~11.3 mg, respectively) in C. albicans (Figure 3a), whereas biomass weights with SU medium significantly increased (~7.7 mg) compared to that with Spider medium (~2.3 mg) in C. tropicalis (Figure 3a). Consistent with a previous report, EFG1 is required for biofilm growth in C. tropicalis (Figure 3a), and a mutant was used as a negative control [20]. Biofilm staining using crystal violet also showed similar conclusions, in which more biofilms were stained in SU medium than in Spider medium in the C. tropicalis wild-type strain (Figure 3b).    Figure 3a. Values are the mean ± SD from four replicates. ** p < 0.01. The C. albicans efg1Δ and C. tropicalis efg1Δ strains were used as negative controls.

Effects of Depletion of Each Ingredient in SU on C. tropicalis Biofilm Growth
To understand why SU is able to induce C. tropicalis biofilm formation, we depleted each ingredient and examined its effect on biofilm growth. The results showed that depletion of MgCl2 or KH2PO4 caused a significant reduction in biofilm formation, especially that of the former (Figure 4a). Lack of CaCl2, Na2SO4, Na3C6H5O7, urea, or creatinine resulted in reduced biofilms but without significant effects (Figure 4a). Given that the magnesium chloride concentration in Spider medium (0.25 g/L) is much lower than that in SU medium (0.65 g/L), we further tested whether increases in the magnesium concentration level in Spider medium to the same level as that in SU were able to promote biofilm growth. As shown in Figure 4b, adding exogenous magnesium did not significantly induce biofilm development, suggesting that it could be a combinational effect of SU in biofilm induction. Furthermore, in order to determine whether growth defects occur in SU medium without any magnesium, thereby causing a severe deficiency in biofilm formation, growth curves were Values are the mean ± SD from four replicates. ** p < 0.01. The C. albicans efg1∆ and C. tropicalis efg1∆ strains were used as negative controls.

Effects of Depletion of Each Ingredient in SU on C. tropicalis Biofilm Growth
To understand why SU is able to induce C. tropicalis biofilm formation, we depleted each ingredient and examined its effect on biofilm growth. The results showed that depletion of MgCl 2 or KH 2 PO 4 caused a significant reduction in biofilm formation, especially that of the former (Figure 4a). Lack of CaCl 2 , Na 2 SO 4 , Na 3 C 6 H 5 O 7 , urea, or creatinine resulted in reduced biofilms but without significant effects (Figure 4a). Given that the magnesium chloride concentration in Spider medium (0.25 g/L) is much lower than that in SU medium (0.65 g/L), we further tested whether increases in the magnesium concentration level in Spider medium to the same level as that in SU were able to promote biofilm growth. As shown in Figure 4b, adding exogenous magnesium did not significantly induce biofilm development, suggesting that it could be a combinational effect of SU in biofilm induction. Furthermore, in order to determine whether growth defects occur in SU medium without any magnesium, thereby causing a severe deficiency in biofilm formation, growth curves were analyzed. The results showed that magnesium depletion caused a striking effect on cell growth, indicating that the presence of magnesium is necessary in proliferation, leading to an impact on biofilm formation (Figure 4c).

ROB1 is not Required for Biofilm Development in C. tropicalis
In C. albicans, six transcription factors, Bcr1, Brg1, Efg1, Tec1, Rob1, and Ndt80, form a complex regulatory circuit controlling biofilm formation. Loss of each gene causes dramatic biofilm defects [16,21]. Genomic sequence data have shown that the six biofilm-associated genes are also present in C. tropicalis [20,26,31]. Six transcription factor amino acid sequences of both C. albicans and C. tropicalis obtained from the Candida Genome Database and UniProt websites were aligned, and alignments were carried out using EMBOSS Pairwise Sequence Alignment. Results showed that C. tropicalis Bcr1, Brg1, Efg1, Tec1, Rob1, and Ndt80 share 53.4%, 44%, 55.4%, 61.4%, 35.9%, and 72.1% identities with each respective homolog in C. albicans (Table 2). CtBrg1 and CtRob1 exhibited much less identity with CaBrg1 and CaRob1 than the others. To determine whether the homologs of major transcriptional factors play crucial roles in biofilm development in C. tropicalis, each C. tropicalis homologous transcriptional gene was deleted. Compared with the wild-type strain (MYA3404), five mutant strains of each gene exhibited a significant reduction in biofilm biomasses (Figure 5a). rob1∆ showed no effect and no statistically significant difference compared with the wild-type strain. Crystal violet assays also revealed results similar to the previous results on silicone (Figure 5b). Furthermore, reintroduction of a functional gene in each mutant but not rob1∆ restored or partially restored biofilm formation in C. tropicalis (Figure 5c). analyzed. The results showed that magnesium depletion caused a striking effect on cell growth, indicating that the presence of magnesium is necessary in proliferation, leading to an impact on biofilm formation (Figure 4c).

ROB1 is not Required for Biofilm Development in C. tropicalis
In C. albicans, six transcription factors, Bcr1, Brg1, Efg1, Tec1, Rob1, and Ndt80, form a complex regulatory circuit controlling biofilm formation. Loss of each gene causes dramatic biofilm defects [16,21]. Genomic sequence data have shown that the six biofilm-associated genes are also present in C. tropicalis [20,26,31]. Six transcription factor amino acid sequences of both C. albicans and C. tropicalis obtained from the Candida Genome Database and UniProt websites were aligned, and alignments were carried out using EMBOSS Pairwise Sequence Alignment. Results showed that C. tropicalis Bcr1, Brg1, Efg1, Tec1, Rob1, and Ndt80 share 53.4%, 44%, 55.4%, 61.4%, 35.9%, and 72.1% identities with each respective homolog in C. albicans (Table 2). CtBrg1 and CtRob1 exhibited much less identity with CaBrg1 and CaRob1 than the others. To determine whether the homologs of major transcriptional factors play crucial roles in biofilm development in C. tropicalis, each C. tropicalis homologous transcriptional gene was deleted. Compared with the wild-type strain (MYA3404), five mutant strains of each gene exhibited a significant reduction in biofilm biomasses (Figure 5a). rob1Δ showed no effect and no statistically significant difference compared with the wild-type strain. Crystal violet assays also revealed results similar to the previous results on silicone (Figure 5b). Furthermore, reintroduction of a functional gene in each mutant but not rob1Δ restored or partially restored biofilm formation in C. tropicalis (Figure 5c). Values are the mean ± SD from five replicates. * p < 0.05; *** p < 0.001. (c) Growth curves of C. tropicalis strains at 30 • C in SU with or without additional MgCl 2 showed the requirement of magnesium for C. tropicalis proliferation. Growth rates were monitored every 2 h using a Biowave density meter. 3.5. Introduction of the C. tropicalis ROB1 into C. albicans Rob1δ Was Not Able to Restore Biofilm Growth in C. albicans To further elucidate the role of C. tropicalis ROB1 (CtROB1) in biofilm formation, we transformed CtROB1 into the C. albicans rob1∆ mutant strains. As shown in Figure 6, CtROB1 in the C. albicans rob1∆ could not promote biofilm development. CtBCR1 and CtEFG1 were selected as the parallel control groups, as deletion of any one gene caused severe impairment of biofilm growth in C. tropicalis ( Figure 5). Results showed that reintroduction of CtBCR1 and CtEFG1 into C. albicans bcr1∆ and efg1∆ strains, respectively, restored biofilm development in C. albicans ( Figure 6). These findings demonstrated that CtROB1 does not play an important role in biofilm formation in C. tropicalis. Furthermore, the results also suggest that, although C. tropicalis and C. albicans are evolutionarily closely related, characteristics of biofilm regulatory circuits of the two species are somehow different.

Introduction of the C. tropicalis ROB1 into C. albicans Rob1δ Was Not Able to Restore Biofilm Growth in C. albicans
To further elucidate the role of C. tropicalis ROB1 (CtROB1) in biofilm formation, we transformed CtROB1 into the C. albicans rob1Δ mutant strains. As shown in Figure 6, CtROB1 in the C. albicans rob1Δ could not promote biofilm development. CtBCR1 and CtEFG1 were selected as the parallel control groups, as deletion of any one gene caused severe impairment of biofilm growth in C. tropicalis ( Figure 5). Results showed that reintroduction of CtBCR1 and CtEFG1 into C. albicans bcr1Δ and efg1Δ strains, respectively, restored biofilm development in C. albicans ( Figure 6). These findings demonstrated that CtROB1 does not play an important role in biofilm formation in C. tropicalis. Furthermore, the results also suggest that, although C. tropicalis and C. albicans are evolutionarily closely related, characteristics of biofilm regulatory circuits of the two species are somehow different.

Discussion
The pathogenicity of chronic and recurrent infections has been associated with the production of biofilms for C. albicans and other non-albicans Candida species [12,15,22]. It is known that different culture conditions affect biofilm growth in several Candida species [19,22,25]. In this study, we established a reliable method for C. tropicalis biofilm development with a silicone-based platform. Three media, RPMI 1640, YNB supplemented with 1% glucose, and SU, as an adhesion medium

Discussion
The pathogenicity of chronic and recurrent infections has been associated with the production of biofilms for C. albicans and other non-albicans Candida species [12,15,22]. It is known that different culture conditions affect biofilm growth in several Candida species [19,22,25]. In this study, we established a reliable method for C. tropicalis biofilm development with a silicone-based platform. Three media, RPMI 1640, YNB supplemented with 1% glucose, and SU, as an adhesion medium were able to induce strong biofilm growth of C. tropicalis. Four media, Spider, Lee's glucose, SD, and SCD supplemented with 50% serum, produced no or less biofilms. Spider medium contains nutrient broth (beef extract and peptone) to supply amino acids, peptides, and proteins [32]. Similarly, Lee's glucose and SCD serum media contain a variety of amino acids [22,25,32]. SD medium contains fewer amino acids [22,25,32]. Interestingly, vitamins and inorganic salts are not the major constituents in these media. By contrast, both YNB and RPMI 1640 are rich in vitamins and inorganic salts, although RPMI 1640 also contains a variety of amino acids [22,25,32]. SU medium contains urea, creatinine, and many inorganic salts [19,24]. This fact implies that vitamins and inorganic salts are important for C. tropicalis adhesion and biofilm growth on silicone. Moreover, the glucose contents of RPMI 1640 (2 g/L), YNB (9 g/L), and SU (3 g/L) media are lower than those of Spider (10 g/L), Lee's glucose (12.5 g/L), SD (20 g/L), and SCD + serum (20 g/L) media. High glucose content in a medium may support and promote planktonic growth rather than biofilm development in C. tropicalis. Interestingly, the compositions of SU are relatively simple, but this medium induced C. tropicalis biofilm formation the most. This result might explain why Candida species favor urinary tract infection [33].
After analysis of the effect of each ingredient in SU on biofilm growth, we found that depletion of magnesium exhibited the greatest impact on cell proliferation, thereby affecting biofilm formation. Metal ions are important elements for prokaryotic and eukaryotic cells and play crucial roles in numerous enzymatic reactions to maintain physiological function, including DNA replication, transcription, and biosynthesis of precursors [34,35], but they can also be toxic to living organisms if present in excess [36]. Furthermore, several reports have shown that cation concentrations influence extracellular product formation and biofilm-associated growth in microorganisms. For example, calcium and magnesium cations can enhance Pseudomonas aeruginosa adherence ability [37]. Increases in Mg 2+ concentration promote cell attachment and biofilm formation in both Staphylococcus aureus and Pseudomonas fluorescens [38,39]. Calcium increases the amount of extracellular matrix composition associated with Pseudoalteromonas sp. biofilms [40]. However, 50 mM and higher Mg 2+ concentrations significantly inhibit biofilm formation in Bacillus species [41]. The adherence ability of Staphylococcus epidermidis is enhanced in low concentrations of magnesium [42]. In C. parapsilosis, an increase in Mn 2+ , with a maximum at 2 mM, significantly induced biofilm formation [43]. Moreover, different metal ions can suppress or enhance C. albicans and C. tropicalis biofilm formation. For example, Co 2+ , Cu 2+ , Ag + , Cd 2+ , Hg 2+ , Pb 2+ , AsO 2 − , and SeO 3 2− inhibit hyphal formation in biofilms of both Candida species, whereas CrO 4 2− triggers a transition to the hyphal cell morphotype in C. tropicalis biofilms [44].
These data indicate that the effects of metal ions on microorganisms are highly dependent on the different types of added metal ions and microbial species. Spider medium, containing fewer magnesium ions than SU medium, is often used for C. albicans biofilm assays, but it fails to induce C. tropicalis adhesion and biofilm formation. The addition of exogenous magnesium to the Spider medium resulted in only very slight, nonsignificant increases in biomass dry weights. These results indicated that C. albicans and C. tropicalis displayed strikingly variable heterogeneous growth, adhesion, and biofilm-forming potential in different growth media. However, the mechanisms of why SU is able to induce C. albicans and C. tropicalis biofilm formation remain unknown. It is possible that the outcome results from the fact that a number of metal ions, urea, and creatinine have minor effects on biofilm formation, contributing to the impact.
Mechanisms underlying biofilm formation in C. albicans are controlled by a central regulatory network with six transcriptional factors: Bcr1, Brg1, Efg1, Tec1, Rob1, and Ndt80 [16]. However, studies regarding biofilm formation mechanisms are very limited in non-albicans Candida species, and biofilm formation regulated by the circuit of six transcription factors may vary in Candida species. For example, in C. parapsilosis, only EFG1 and BCR1 deletion strains exhibited similar effects on biofilm formation as those observed in C. albicans, whereas BRG1 and TEC1 do not play critical roles in biofilm growth in C. parapsilosis [45]. Furthermore, in contrast to in C. albicans, four transcription factor genes, CPH2, UME6, CZF1, and GZF3, were characterized for their roles in biofilm formation and are unique to C. parapsilosis [45]. These six homologous genes have also been identified in C. tropicalis [20,26]. Interestingly, the degree of sequence identity of two species corresponds to biofilm results, in which the lower identity causes less effect in biofilm formation in C. tropicalis mutant strains. Both C. tropicalis brg1∆ and rob1∆ strains formed more biofilms compared to the other mutant strains, although brg1∆ produced reduced biofilm significantly. Furthermore, expression of CtROB1 in the C. albicans rob1∆ was not able to recover biofilms. These findings suggest that, although the functions of five biofilm-associated regulators in both C. albicans and C. tropicalis are relatively conserved, mechanisms regarding biofilm formation in C. tropicalis are distinct from those of its relative C. albicans. Furthermore, it will be worthwhile to explore the transcriptional expression profile between planktonic cells and biofilm cells using an SU-induced biofilm platform to understand the C. tropicalis biofilm regulatory network.