The Lack of SNARE Protein Homolog Syn8 Influences Biofilm Formation of Candida glabrata

Biofilm formation of Candida species is considered to be a pathogenic factor of host infection. Since biofilm formation of Candida glabrata has not been as well studied as that of Candida albicans, we performed genetic screening of C. glabrata, and three candidate genes associated with biofilm formation were identified. Candida glabrata SYN8 (CAGL0H06325g) was selected as the most induced gene in biofilm cells for further research. Our results indicated that the syn8Δ mutant was defective not only in biofilm metabolic activity but also in biofilm morphological structure and biomass. Deletion of SYN8 seemed to have no effect on extracellular matrix production, but it led to a notable decrease in adhesion ability during biofilm formation, which may be linked to the repression of two adhesin genes, EPA10 and EPA22. Furthermore, hypersensitivity to hygromycin B and various ions in addition to the abnormal vacuolar morphology in the syn8Δ mutant suggested that active vacuolar function is required for biofilm formation of C. glabrata. These findings enhance our understanding of biofilm formation in this fungus and provide information for the development of future clinical treatments.


INTRODUCTION
Candida glabrata, identified as a non-albicans Candida (NAC) species, is one of the common causes of systemic candidiasis and shows resistance to most azole drugs (Rodrigues et al., 2014;Pristov and Ghannoum, 2019). However, studies on the mechanisms of virulence in C. glabrata are far fewer than those in C. albicans. The biofilm of Candida species is known to be a pathogenic factor in human infections and difficult to eradicate in healthcare settings (Douglas, 2003;Eix and Nett, 2020). In C. glabrata, the biofilm is composed of only yeast-shaped cell clusters packed in a multilayer, which is quite different from biofilms of C. albicans with their elongated filamentous forms (Brunke and Hube, 2013). Because C. glabrata lacks filamentous growth, it is important to investigate its own distinctive biofilm pathogenicity.
Biofilm formation is a complex process that includes adhesion, colonization, extracellular matrix production, biofilm maturation, and dispersal (Ramage et al., 2009). The first step is Candida cell adhesion to the surfaces of host cells and biomaterials such as silicone. Therefore, many previous studies have investigated the roles of adhesins at the cell surface in biofilm formation. The EPA (epithelial adhesin) gene family is a major group of adhesins in C. glabrata. Because the C. glabrata haploid genome is more closely related to Saccharomyces cerevisiae than C. albicans, this large EPA family is similar to the FLO gene family of S. cerevisiae, which encodes lectins for flocculation (de Groot et al., 2008). In particular, EPA3 and EPA6 have been indicated as encoding the main adhesins involved in biofilm formation of C. glabrata (Iraqui et al., 2005;Cavalheiro et al., 2018). Furthermore, EPA6 expression is positively regulated by a Yak1p kinase that acts through a subtelomeric silencing pathway and by the chromatin remodeling Swi/Snf complex, whereas it is negatively regulated by the transcription factor Cst6p (Iraqui et al., 2005;Riera et al., 2012). The Swi/Snf nucleosome-remodeling complex interacts with the nucleosomes via the Snf2p and Snf6p components. SNF2 and SNF6 gene disruption mutants also exhibit poor ability to form biofilms (Riera et al., 2012). Pgs1p is a phosphatidylglycerol phosphate synthase, and it has been reported that deletion of the PGS1 gene results in increased cell surface hydrophobicity and decreased biofilm formation of C. glabrata (Bat'ová et al., 2009). In addition to proteins that act on cell wall synthesis related to the adhesion step during biofilm formation, some recent studies have reported that the multidrug resistance transporters Pdr16p and Tpo1_2p contribute to azole resistance and are related to biofilm formation (Culakova et al., 2013;Santos et al., 2017). More recently, Dtr1p, Qdr2p, and Tpo4p drug:H + antiporters were found to be associated with biofilm formation of C. glabrata (Widiasih Widiyanto et al., 2019;Santos et al., 2020). Thus, it seems like a broad range of proteins, which are valuable to study, affect the process of biofilm formation of C. glabrata.
Comprehensive screening of biofilm-related genes in fungi was commonly used the collections of gene mutants (Iraqui et al., 2005;Vandenbosch et al., 2013), and the genetic expression analysis comparing the biofilm form with planktonic cells (García-Sánchez et al., 2004;Nobile et al., 2012). However, these previous studies have mostly used C. albicans and S. cerevisiae rather than C. glabrata. The aim of this study was to identify new genes involved in biofilm formation of C. glabrata. Through a comprehensive screening of 101 C. glabrata mutants, three gene mutants exhibited less than 70% of the metabolic activity compared with the wild type strain. Our study focused on one of the new genes, SYN8 (CAGL0H06325g), which encodes a SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) protein in C. glabrata. Although there have been many studies on the roles of cell wall proteins in the biofilm formation of C. glabrata, the relationship between biofilm formation and SNARE proteins is unclear. This is the first study to investigate the role of the SNARE Syn8p in C. glabrata, and we demonstrated that this gene is involved in normal vacuolar function. In addition, our results suggested that SYN8 influences the biofilm formation of C. glabrata.

Strains and Media
The C. glabrata strains used in this study are listed in Table 1 and Supplementary Table 1. All yeast strains were typically grown in YPD medium (1% Bacto yeast extract, 2%

Plasmid Construction
The C. glabrata mutant library was constructed using the method described previously (Ueno et al., 2007), and the information of gene sequence was provided by Candida Genome Database (Koszul et al., 2003). The gene deletion was carried out in the parental strain KUE100. The target gene was replaced by a DNA replacement cassette including the CgHIS3 gene through homologous recombination. The pHIS906 plasmid including CgHIS3 was used as a template ( Figure 1A). Recombination loci and gene deletion were verified by PCR.
To construct the SYN8 reintegrated strain from the syn8 mutant, C. glabrata wild type genomic DNA was extracted from the CBS138 strain and used as the template to amplify the SYN8 ORF with a 500 bp promoter and 200 bp terminator. The primers pSyn8compF and pSyn8compR were designed to amplify the SYN8 (1.5 kb) fragments. Next, the plasmid pZeoi_comp606 and the amplified 1.5 kb fragment were digested, respectively, and then ligated together to construct pZeoi_SYN8 ( Figure 1B). The sequence of the SYN8 gene was confirmed by a DNA sequencer (Open Research Facilities for Life Science and Technology, Tokyo Institute of Technology). The primers pChr606F1 and pChr606R1 were used to amplify the cassette from pZeoi_SYN8, and the purified DNA product was transformed into the noncoding region on chromosome F position 605,901-606,015 of the mutant syn8 . Successful integration at the right position on the transformant genome was confirmed by the primers pZeoORFcheckF1 and pchr606check. All primer sequences used in this experiment are shown in Table 2 and designed by the software Primer 3.

Transformation of C. glabrata
After constructing the integration cassettes, the cassettes were introduced into the C. glabrata syn8 strain based on a previous method with slight modification (Ueno et al., 2011). The strain colonies were incubated in YPD liquid medium and cultured overnight with shaking at 37 • C. The cultured cells were resuspended in 10 mL fresh YPD medium and grown from an OD 600 of 0.4-1.0 at 37 • C. These cells were resuspended in 10 mL of 0.15 M lithium acetate dissolved in TE buffer (LiOAc/TE) and shaken lightly for 1 h at 37 • C. Then the cells were again harvested and resuspended in 400 µL of 0.15 M LiOAc. Sixty microliters of the cell suspension were supplemented with 20 µg single-stranded carrier DNA and 5-10 µg integration cassette DNA solution and mixed gently in a new 1.5 mL tube. The cell suspension was incubated at 37 • C with rotation for 37 min and then mixed completely with 120 µL of 52.5% polyethylene glycol 4000, 0.15 M LiOAc by pipetting. The cells were incubated again for 45 min at 37 • C and heat-shocked at 42 • C for 15 min. Then, the cells were harvested and incubated in 1 mL of fresh YPD liquid medium for 12 or 24 h at 37 • C with shaking for recovery. After that, cells were spread on YPD plates containing zeocin (50-200 µg/mL) and incubated at 37 • C for 1-2 days.

Biofilm Formation and Plate Siliconization
C. glabrata biofilms were formed according to an in vitro method using a 96-well flat-bottom polystyrene plate (IWAKI, Japan) (Widiasih Widiyanto et al., 2019). Yeast cells were cultured overnight at 37 • C. Cells were harvested by centrifugation and washed twice in 100 µL sterile phosphate-buffered saline buffer (PBS, 137 mmol/L NaCl, 2.7 mmol/L KCl, 10 mmol/L Na 2 HPO 4 , 2 mmol/L KH 2 PO 4 , pH 7.2). Then cells were resuspended in fresh medium and standardized to a concentration of 10 7 cells/mL. Each well of the plate was filled with a 100 µL aliquot of cell suspension, and the plates were incubated at 37 • C for 1.5 h to allow the cells to adhere to the surfaces of the wells. Following the adhesion phase, the cell suspensions were removed, and the wells were washed with PBS buffer to remove non-adherent cells.
Fresh medium (100 µL) was added to each well, and then the plate was incubated at 37 • C for 24 h. The siliconized plates were also used to compare the influence of the silicone material on biofilm formation. Three hundred microliters of Sigmacote (Sigma-Aldrich, United States) were added into the well of a plate and distributed over the well according to the product information. Because the reaction was almost instantaneous, the rest of the Sigmacote was quickly removed. After all of the wells of one plate were treated, the whole plate was dried in a hood with UV sterilization. The silicone is covalently bound to the well.

XTT Reduction Assay
2,3-Bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenylamino) carbonyl]-2H-tetrazolium-5-carboxanilide (XTT) reduction assay was used to measure the metabolic activity of biofilms of C. glabrata as described previously (Chen et al., 2017). Briefly, after washing the biofilm with 100 µL PBS buffer, a 100 µL aliquot of XTT-menadione was added into each well. The plate was then incubated for 2 h at 37 • C, and the colorimetric change was measured at 490 nm by a microtiter plate reader (VarioskanLUX, ThermoFisher, Japan). Negative controls were the unseeded wells. The final result for each strain was confirmed at least three independent times.

Observation of Biofilms by SEM
To observe the microstructures of C. glabrata biofilms, sterile 3-mm cubic silicone sponges (AS ONE Corp., Japan) were prepared and put into a 96-well plate before cell cultivation. Aliquots (200 µL) of precultured cell suspensions were added in the plate, and the biofilm was allowed to form subsequently on the silicone sponges. These samples were treated as described before (Chen et al., 2017). The biofilm microstructures of C. glabrata were observed by a desktop scanning electron microscope, Phenom TM -Pro-X (Phenom-World, Netherlands).

Biofilm Dry Weight and Matrix Analysis
Biofilms for matrix collection were formed in 12-well polystyrene microtiter plates (IWAKI, Japan). Three-milliliter aliquots of cell suspensions that were diluted to 1 × 10 7 cells/mL were filled in each well. After incubation at 37 • C for 24 h, the matrix of each biofilm was collected using a previously described method with slight modification (Silva et al., 2009;Taff et al., 2012). Biofilms were scraped from the wells and resuspended with 5 mL distilled water in tubes. The tubes were sonicated for 10 min, vortexed vigorously for 5 min, and centrifuged at 5,000 × g for 10 min at 4 • C. The supernatant was transferred into new tubes to separate cell pellets from soluble matrix material and stored at −20 • C. To measure the biofilm dry weight, the remaining pellets were dried at 60 • C until a constant weight. The supernatant samples were used to determine the contents of proteins and total carbohydrates. The protein content of each biofilm matrix was measured by using Protein Assay CBB Solution (5× ) (Nacalai Tesque, Japan). Bovine serum albumin (BSA) (2.5-25 µg/mL) was used to make the standard curve for the protein test. The absorbance was read at 595 nm. The total carbohydrate content of each biofilm matrix was detected based on the phenol-sulfuric acid method (Dubois et al., 1956). Briefly, 1 mL sample was mixed well with 250 µL 80% phenol (wt/wt in H 2 O) and added rapidly with 2.5 mL H 2 SO 4 . Finally, after the tubes cooled down to room temperature, the samples were mixed well and the absorbance was read at 490 nm. Glucose (5-500 µg/mL) was used to make the standard curve for the total carbohydrate test.

Real-Time PCR
The planktonic cells were cultured in CSM medium at 37 • C. The culture was diluted to an OD 600 of 0.2 in 50 mL fresh CSM medium before being grown to OD 600 of 1∼2. The biofilm cells were induced to form in 12-well plates for 24 h as described above. Total RNA was extracted from unicellular cells and biofilm cells using the glass bead lysis method. These RNAs were used as templates to synthesize first strand cDNA using the ReverTra Ace qPCR RT Master Mix with gDNA Remover kit (TOYOBO, Japan). Real-time PCR was performed by using the THUNDERBIRD SYBR qPCR Mix kit (TOYOBO, Japan) and gene expression levels were determined by using C t method. The gene expression values were normalized by comparison to the expression of a housekeeping gene, CgACT1 (Pfaffl, 2001). All the primers for real-time PCR reactions were listed in Table 3.

Susceptibility to Drugs and Ions
Cells were cultured in 5 mL YPD medium overnight at 37 • C. The overnight cultures were then diluted to an OD 600 of 0.1, which was used as a starting point for 10-fold serial dilutions. Three microliters of each serial dilution were spotted onto YPD plates containing drugs and ions (Banuelos et al., 2010;Palanisamy et al., 2010). The final concentrations of each drug were 0.01%

Antifungal Susceptibility of Unicellular Cells by Broth Dilution
Minimum inhibitory concentration (MIC) assays were prepared based on EUCAST guidelines (Rodríguez-Tudela et al., 2003). Briefly, cells were cultured overnight in CSM medium, washed twice by PBS buffer, and resuspended in RPMI medium to a concentration of 10 4 cells/mL. Volumes of 100 µL of cells including hygromycin B were filled in 96-well microtiter plate wells. The concentration range of the 2-fold drug dilutions was from 7.8125 to 500 µg/mL. The plates were incubated at 37 • C for 24 h and then read by a microtiter plate reader at 530 nm.

Antifungal Susceptibility of Biofilm Cells by XTT Assay
Sessile Minimum Inhibitory Concentration (SMIC) of biofilm cells was tested as previously published (Ramage et al., 2001). Biofilms were formed as described above. After 24 h, the suspension was discarded, and wells were washed twice by 100 µL PBS buffer. In order to avoid disruptions of biofilms, different concentrations of 2-fold antifungal agent dilutions were prepared separately in medium (62.5-1,000 µg/mL), and 100 µL amounts of each antifungal concentration were added into the wells. Controls were prepared with biofilm wells that were untreated with drugs or unseeded wells with medium only. The plates were incubated at 37 • C for 24 h again. After 24 h, the plates were carefully washed with 100 µL PBS. To determine XTT activity, 100 µL of a 1 mg/mL solution of XTT-menadione was put into each well and the plates incubated for 2 h. Absorbance at 492 nm was measured by a microtiter plate reader. The lowest concentrations associated with a 50 or 90% reduction in absorption compared with the untreated biofilm wells were reported as the MIC 50 or MIC 90 .

RESULTS
A comprehensive screening of C. glabrata mutants was performed by using our C. glabrata mutant library. The null mutants of 101 genes that encode proteins including signal peptides or transmembrane domains were selected and screened for their biofilm formation abilities. The biofilm formation abilities of 101 mutants were compared with that of the reference strain ( Supplementary Figure 1 and Supplementary Table 1). In comparison with the wild type, three gene mutants of C. glabrata showed significant reductions in biofilm formation, and the metabolic activities of these mutant biofilms were less than 70% of that of the wild type strain. The expressions of all three of these genes in C. glabrata during biofilm formation were detected as upregulated compared with those in planktonic cell conditions (Supplementary Figure 2). In particular, the expression levels of two genes in biofilm formation were more than two times higher than in planktonic cells. From a search of the Candida genome database, we found that one gene is a homolog of the SNARE protein family of S. cerevisiae. Since this gene showed the most drastic induction in the biofilm cells, we focused on this gene SYN8 (CAGL0H06325g) and characterized its deletion as the syn8 mutant. SYN8 in C. glabrata is a homolog of SYN8 in S. cerevisiae and of CR_03140C_A/SYN8 in C. albicans. In this study, the protein sequence of C. glabrata SYN8 was aligned with those of S. cerevisiae SYN8 and C. albicans SYN8, and the BLAST results showed that C. glabrata Syn8p shares 40% identity with S. cerevisiae Syn8p and 33% identity with C. albicans Syn8p.
Next, the growth rate of the C. glabrata syn8 strain was measured and compared with those of the wild type and a SYN8 reintegrated strain. The syn8 mutant grew similarly to the wild type and reintegrated strains with a 24 h incubation in CSM medium, which was also used in the biofilm formation experiments (Supplementary Figure 3). After that, the biofilm formation of the syn8 mutant was compared to those of the wild type strain and the reintegration SYN8 strain. On the non-siliconized plates made from polystyrene, the metabolic activity of the syn8 mutant cells was significantly lower than those of the wild type strain and the SYN8 reintegrated strain when measured by XTT assay. In the siliconized plates (silicone coated polystyrene), the biofilm formation of the syn8 mutant showed almost the same proportional reduction as that on the non-siliconized plates, while the reintegration SYN8 strain kept the same levels of biofilm formation compared to the wild type strain (Figure 2A). From these results, we used the non-siliconized plates in further experiments. To further compare the quantities of mature biofilm cells among the three strains, the biomass of the biofilms was measured. The dry weights of biofilms of the syn8 mutant were almost half as much as those of the wild type strain, while FIGURE 2 | Effect of siliconized plates on biofilm formation. (A) The metabolic activities of biofilms of the wild type strain, syn8 mutant, and SYN8 reintegrated strain were detected in non-siliconized plates and siliconized plates, respectively. (B) Biofilms of the wild type strain, syn8 mutant, and SYN8 reintegrated strain were respectively formed in 12-well plates, and the dry weight was measured. The reported values are the means ± SD of three independent experiments. * p < 0.05 with Student's t-test. * vs. WT or SYN8.
Frontiers in Cell and Developmental Biology | www.frontiersin.org the reintegration SYN8 strain showed no decrease in biofilm biomass ( Figure 2B). The biofilm structures of these three strains on silicone sponges were observed by SEM. The SEM images showed that biofilms of the wild type strain extend widely with many cells, while few cells aggregated in the biofilm of the syn8 mutant (Figure 3). The reintegration SYN8 strain recovered this phenotype of the mutant strain, indicating that the SYN8 gene is involved in the development of biofilms in C. glabrata. The reductions in metabolic activity, biomass, and biofilm development in the syn8 mutant suggested that the SYN8 gene plays an important role in biofilm formation of C. glabrata.
Matrix analysis was performed to detect the carbohydrate and protein components in the matrix of biofilms. Table 4 shows the yield of total carbohydrate and protein separated from biofilm matrix formed by the wild type, the syn8 mutant, and the reintegration strain. The total carbohydrate and protein in the biofilm matrix formed by the syn8 mutant was much less than in the wild type and reintegration strains. However, when the amounts of carbohydrate and protein were normalized for the dry weight of the biofilm cells, the values showed no significant difference. This suggested that deletion of SYN8 had no effect on the proportion of biofilm matrix components in C. glabrata.
The first step in biofilm formation is cell adherence to the surface. The adherent activity of the syn8 mutant during biofilm formation was also analyzed. After an initial 1.5 h of biofilm formation in siliconized plates, the non-adherent FIGURE 3 | Morphology of 24 h biofilms. Biofilms of the wild type strain, syn8 mutant, and SYN8 reintegrated strain were formed on silicone sponges in CSM medium for 24 h. Biofilms of these three strains were visualized by SEM at 300×, 1,000×, and 5,000× magnification. Scale bars indicate 150 µm in 300× images, 50 µm in 1,000× images and 10 µm in 5,000× images. The reported values are the means ± SD of three independent experiments. cells were washed away by PBS, and the metabolic activities of cells adherent to the surface of the plate bottom were measured by XTT assay. The syn8 mutant showed around 70% metabolic activity compared with the wild type and reintegration strains ( Figure 4A). When observing the adhesion step by SEM imaging, the results also showed that there were far fewer cells of the syn8 mutant attached on the surface of the silicone compared with the wild type and reintegration strains ( Figure 4B). Therefore, normal adhesion of biofilms seems to require Syn8p. Because the syn8 mutant was observed to show poor adhesion, four adhesin genes, EPA1, EPA6, EPA10, and EPA22, were selected to investigate their transcriptional expression levels in the syn8 mutant and the control strains under planktonic cell growth and biofilm formation conditions. There were no significant differences in the transcriptional expression levels of EPA1 and EPA6 genes under either condition. However, EPA10 and EPA22 showed decreased expression in the syn8 mutant under both conditions compared with the wild type control (Figures 5A,B).
Although we found that the SYN8 gene contributes to biofilm formation and surface adhesion, it remained unclear whether this gene was related to stress resistance in biofilms of C. glabrata. Therefore, the growth of the syn8 mutant in the presence of several antimicrobial reagents was detected by spot dilution assay. Compared to the wild type and SYN8 reintegrated strains, the syn8 mutant showed significant defects in growth with SDS and hygromycin B. However, deletion of SYN8 seemed to show no change in susceptibility to calcofluor white and Congo red, which interfere with cell wall assembly (Figure 6).
Since these antimicrobial treatments show poor activity against biofilm cells when compared with planktonic cells, the MIC of hygromycin B to planktonic cells and the SMIC of hygromycin B to biofilm cells were measured. The MIC 50 and MIC 90 of planktonic cells of the syn8 mutant were respectively 7.8125 and 62.5 µg/mL for hygromycin B, values that were much lower than those of the wild type strain (62.5 and 250 µg/mL) and SYN8 reintegrated strain (125 and 250 µg/mL). The SMIC 50 and SMIC 90 of the syn8 mutant biofilm cells were 125 and 1,000 µg/mL, respectively, which were much higher than those of its own planktonic cells. However, the SMIC 50 of the wild type strain and the integrated strain were 500 and 1,000 µg/mL, respectively, and the SMIC 90 of these strains was above 1,000 µg/mL (Table 5).
To synthesize mature proteins in cells normally, vacuolar functions such as vacuolar transport are necessary, and Syn8p is thought to be related to vacuolar function. To check the  5 h) of the wild type, syn8 mutant, and SYN8 reintegrated strains were tested by XTT reduction assay. The reported values are the means ± SD of three independent experiments. * p < 0.05 with Student's t-test, * vs. WT or SYN8. (B) The morphology of adhesion cells of the wild type, syn8 mutant, and SYN8 reintegrated strains were visualized by SEM at 300 × magnification. Scale bars indicate 150 µm in 300× images.
FIGURE 5 | The transcriptional expression levels of EPA1, EPA6, EPA10, and EPA22 in C. glabrata under planktonic cell growth conditions (A) and biofilm formation conditions (B). The gene expression values were normalized by comparison to the expression of a housekeeping gene, CgACT1. 100% to the control group of wild type strain. Reported values indicate the means ± SD of at least three independent experiments. * p < 0.05 with Student's t-test.
FIGURE 6 | Susceptibility to antifungals by spot-dilution assays. Overnight cultures were diluted in 10-fold serial dilutions and spotted onto YPD plates containing 0.01% SDS, 70 µg/mL hygromycin B, 80 µg/mL calcofluor white, and 100 µg/mL Congo red, respectively. vacuolar function of the syn8 mutant, it was assayed for growth under various types of ionic stress. The syn8 mutant showed higher sensitivities to ZnCl 2 , MnCl 2 , and NaCl than the wild type strain (Supplementary Figure 4). In order to examine vacuole morphology, the fluorescent dye FM4-64 was used to stain vacuolar membranes and was viewed by CLSM. FM4-64 is a lipophilic styryl dye. Both the wild type and SYN8 reintegrated strains showed clear circle vacuolar membranes that were stained by FM4-64. But syn8 mutant cells did not have a clear vacuole shape and looked to have large numbers of fragmented vacuoles (Figure 7). This result suggested that deletion of the C. glabrata SYN8 gene resulted in defective vacuole morphology.

DISCUSSION
Syn8 protein in S. cerevisiae has been identified as the SNARE protein syntaxin 8, which plays a role in transport between the Golgi and prevacuolar compartments (PVC) (Lewis and Pelham, 2002). Previous findings have indicated that Syn8p may be a constituent of a SNARE complex with Vti1p, Pep12p, and Ykt6/Nyv1 proteins (Fischer Von Mollard and Stevens, 1999;Kweon et al., 2003). Some studies on the relationship between biofilm formation and SNARE proteins in C. albicans and S. cerevisiae have been previously conducted. Pep12, which was mentioned above, has been reported to act as a t-snare protein and is classified into the Qa-/Syntaxin family (Lewis and Pelham, 2002). It has been reported to be involved in vacuolar transport and endocytosis in both S. cerevisiae and C. albicans (Kweon et al., 2003;Palanisamy et al., 2010). During biofilm formation of the C. albicans pep12 null mutant, there are notable defects in biofilm mass and biofilm integrity (Palanisamy et al., 2010). Another v-snare protein, Btn2, has also been demonstrated to play a role in biofilm formation in S. cerevisiae flor strains. Btn2p facilitates the trafficking of specific proteins from the late endosome to the Golgi apparatus. Furthermore, the deletion of BTN2 affects biofilm formation due to alterations of the transcriptional level of the FLO11 gene, a critical gene for biofilm formation in S. cerevisiae (Espinazo-Romeu et al., 2008). Our study is the first to demonstrate a role for a SNARE protein in biofilm formation of C. glabrata.
Candida biofilms have been reported to exist frequently in the bodies of patients with silicone implant devices and are highly resistant to antifungal drugs (Pristov and Ghannoum, 2019). Medical devices made of silicone materials include urinary catheters, ventriculoperitoneal shunts, and voice prosthesis (Kojic and Darouiche, 2004). Therefore, not only commercial microtiter plates but also siliconized plates were used for biofilm formation of C. glabrata in this study. The syn8 mutant decreased the metabolic activity of biofilms in both plate types. Furthermore, the syn8 mutant seemed to be unable to form layers of dense biofilm structures and construct the same biofilm biomass as the wild type. Our in vitro model results strongly suggest that SYN8 is required for proper C. glabrata biofilm formation. Further work will be need to elucidate the function of this gene in biofilm model in vivo.
Mature biofilm structures consist of yeast cell layers and extracellular matrix, which mainly includes carbohydrates and proteins (Silva et al., 2009). The protein and carbohydrate amounts were also reduced in the syn8 mutant, although the absence of the SYN8 gene did not influence the ratio of extracellular matrix components to biofilm biomass. Moreover, deletion of the SYN8 gene in C. glabrata resulted in a decrease in the adhesion to the surface of the silicone plates. Since adhesion and production of extracellular matrix are significant steps to from biofilms of C. glabrata, these results suggested that SYN8 might affect not only the adhesion ability in the early stage of biofilm formation, but also the extracellular matrix production in the late stage of biofilm formation.
The EPA gene family is a major group of adhesin-like glycosylphosphatidylinositol-cell-wall proteins (GPI-CWPs) of C. glabrata. Epa1p plays a critical role in adherence in C. glabrata. Epa6p has also been shown to play an important role during biofilm formation although it is not involved in adherence to the surface of plates (Iraqui et al., 2005). EPA1, EPA6, and EPA10 share the same homolog in S. cerevisiae, FLO1, while the EPA22 homolog in S. cerevisiae is FLO10. The Flo protein family in S. cerevisiae has been reported to mediate yeast flocculation, adherence, and biofilm formation (Verstrepen and Klis, 2006). In order to clarify the relationship between EPA gene expression and SYN8 disruption in C. glabrata, the transcriptional expressions of EPA1, EPA6, EPA10, and EPA22 during planktonic cell growth and biofilm formation were analyzed in the syn8 mutant. Since the expression levels of both EPA10 and EPA 22 in the syn8 mutant were much lower in biofilm formation than planktonic cell growth, these proteins may be needed for adhesion activity in biofilms of C. glabrata. Further research into these two EPA genes will be required to determine their relationship with the SYN8 gene.
In addition, testing the resistance to several stressors demonstrated that the deletion of SYN8 led to increased susceptibility to hygromycin B especially, not only in planktonic cells but also in biofilm cells. Hygromycin B inhibits translocation of mRNA and tRNA on the ribosome and interferes with decoding fidelity. It is widely used in veterinary medicine and in selection for cell culture (Borovinskaya et al., 2008). Based on previous research on S. cerevisiae, many mutants such as VPS1, VPS34, VPS45, and VPS54 with vacuole trafficking or functional defects show hypersensitivity to hygromycin B (Banuelos et al., 2010;Smaczynska-de Rooij et al., 2010). Moreover, these VPS mutants have decreased growth in the presence of various ions, similar with the syn8 mutant in this study. This could be because vacuoles act as storage vessels for a wide variety of ions, responding to ionic shock and promoting intracellular ion homeostasis. For instance, Na + and K + are major cations in the cytosol, while they are minor components in the vacuolar pools (Rodríguez-Navarro, 2000). Excessive concentrations of Mg 2+ and Zn 2+ can become harmful to cells (Simm et al., 2007). In addition, our CLSM images showed that the vacuole morphologies of the syn8 mutant were abnormal when compared to the wild type strain. In S. cerevisiae, fragmented vacuoles are observed in the VPS54 null mutant (Raymond et al., 1992). From our findings, the syn8 mutant of C. glabrata appeared to have defective vacuolar morphology and the detail study on the role of SYN8 gene in the vacuolar function is expected in future research.
Vps proteins of S. cerevisiae have been reported to be required for the assembly of SNARE complexes during fusion between the trans-Golgi network and PVC. For example, some Vps proteins, together with the vacuolar SNARE proteins, function in the fusion of multiple transport intermediates with the vacuole (Sato et al., 2000). Some C. albicans Vps proteins, like Vps1p, have also been reported to be involved in biofilm formation (Bernardo et al., 2008). Therefore, it has been proposed that SNARE proteins and Vps proteins may act together on the secretory pathway, which plays a role in Candida virulence, such as biofilm formation (Rollenhagen et al., 2020).

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
XC and SK conceived the study and wrote the manuscript. XC performed the experiments. XC and SI collected and analyzed the data. TK and HC gave support. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We thank the Open Research Facilities for Life Science and Technology in the Tokyo Institute of Technology for technical support and equipment and Tokyo Institute of Technology for internal funds to us.
Supplementary Table 1 | Biofilms of mutants in percentages relative to the reference strain (100%).