Bacillus subtilis natto Derivatives Inhibit Enterococcal Biofilm Formation via Restructuring of the Cell Envelope

Enterococcus faecalis is considered a leading cause of hospital-acquired infections. Treatment of these infections has become a major challenge for clinicians because some E. faecalis strains are resistant to multiple clinically used antibiotics. Moreover, the presence of E. faecalis biofilms can make infections with E. faecalis more difficult to eradicate with current antibiotic therapies. Thus, our aim in this study was to investigate the effects of probiotic derivatives against E. faecalis biofilm formation. Bacillus subtilis natto is a probiotic strain isolated from Japanese fermented soybean foods, and its culture fluid potently inhibited adherence to Caco-2 cell monolayers, aggregation, and biofilm production without inhibiting the growth of E. faecalis. An apparent decrease in the thickness of E. faecalis biofilms was observed through confocal laser scanning microscopy. In addition, exopolysaccharide synthesis in E. faecalis biofilms was reduced by B. subtilis natto culture fluid treatment. Carbohydrate composition analysis also showed that carbohydrates in the E. faecalis cell envelope were restructured. Furthermore, transcriptome sequencing revealed that the culture fluid of B. subtilis natto downregulated the transcription of genes involved in the WalK/WalR two-component system, peptidoglycan biosynthesis and membrane glycolipid biosynthesis, which are all crucial for E. faecalis cell envelope synthesis and biofilm formation. Collectively, our work shows that some derivatives present in the culture fluid of B. subtilis natto may be useful for controlling E. faecalis biofilms.


INTRODUCTION
Enterococci, which are Gram-positive bacteria normally present in human gastrointestinal tracts, are the second most common pathogens recovered from catheter-associated infections of the bloodstream and urinary tract and from skin and soft-tissue infections in hospitals in the United States (Paulsen et al., 2003;Arias and Murray, 2012). Among Enterococcus species, Enterococcus faecalis is the primary species responsible for human enterococcal infections (Sievert et al., 2013).
Treatment of E. faecalis infections has become increasingly difficult because of the emergence of E. faecalis strains that are resistant to numerous clinically used antibiotics, such as macrolides; tetracyclines; aminoglycosides; and glycopeptides, including vancomycin, which was previously used as the antibiotic of last resort for enterococcal infections (Murray, 1997;Klevens et al., 2007;Arias et al., 2010;Frieden, 2013). Moreover, E. faecalis has a propensity to transfer antibiotic resistance genes to other bacteria within and across species via pheromone-inducible conjugative plasmid transfer, which facilitates the dissemination of antibiotic resistance (Clewell, 1990;Clewell et al., 2002).
In addition to having intrinsic resistance to multiple antibiotics and the ability to transfer antibiotic resistance via plasmid transfer, E. faecalis can readily form biofilms on a wide range of natural and artificial substrates, such as damaged heart valves, venous catheters, urinary catheters, and indwelling medical devices (Donlan and Costerton, 2002;Fernandez Guerrero et al., 2007). Biofilms are aggregates of microbes that accumulate at a solid-liquid interface and are encased in a self-produced matrix of extracellular polymeric substances (Flemming et al., 2007;Flemming and Wingender, 2010). Since the protective extracellular matrix can decrease the penetration of antibiotics, E. faecalis cells in biofilms can be 10 to 1,000 times more resistant to antibiotics than their planktonic counterparts (Hoyle and Costerton, 1991;Mah and O'Toole, 2001). This trait of enterococcal biofilms markedly reduces the effectiveness of current antibiotic treatments. In addition, enterococcal biofilms have been shown to serve as nidi for bacterial dissemination and as reservoirs for antibiotic resistance genes (Ch'ng et al., 2019). Taken together, the evidence indicates that the presence of E. faecalis biofilms can make infections with E. faecalis more difficult to eradicate. Therefore, there is a demand for novel, safe, and effective methods to inhibit the formation of E. faecalis biofilms.
In addition to discovering and developing new antibiotics, scientists have explored the possibility of preventing and treating gastrointestinal tract infections with probiotics, which are live microorganisms, such as bacteria and yeast, that can provide benefits to the host when administered in adequate amounts (Saarela et al., 2002;Kamada et al., 2013). The use of sporeforming bacteria, mostly of the genus Bacillus, as probiotics has attracted considerable attention from researchers in recent years (Elshaghabee et al., 2017). In comparison to commonly used non-spore-forming probiotic lactic acid bacteria, Bacillus species can form spores under harsh environments. This trait enables them to have higher acid tolerance and better stability during heat processing and low-temperature storage than other bacteria (Elshaghabee et al., 2017). In addition, previous studies (Piewngam et al., 2018;Tazehabadi et al., 2021) have shown that some Bacillus species possess the ability to inhibit the colonization and biofilm formation of pathogens via actions, such as interference with quorum-sensing signals, or production of antimicrobial agents, such as bacteriocin proteins. Piewngam et al. (2018) found that probiotic Bacillus subtilis can produce the lipopeptide fengycin for decolonization of methicillin-resistant Staphylococcus aureus (MRSA) in mouse feces and intestines via interference with S. aureus agr quorum-sensing signaling. Tazehabadi et al. (2021) found that two bacteriocin-producing Bacillus probiotic strains, B. subtilis KATMIRA1933 and Bacillus amyloliquefaciens B-1895, can inhibit the biofilm formation of several strains of the food-borne pathogen Salmonella enterica without killing planktonic S. enterica cells. Collectively, these studies support the idea that probiotic Bacillus species and their derivatives may have the potential to inhibit the biofilm formation ability of other pathogenic bacteria, such as E. faecalis.
B. subtilis natto is a probiotic strain isolated from natto, which is a traditional fermented soybean food in Japan with a long history of consumption (Nishito et al., 2010). In the process of steamed soybean fermentation, B. subtilis natto produces various derivatives, such as extracellular proteases, viscous substance γ-poly-DL-glutamic acid (γ-PGA), and antibiotics (Nishito et al., 2010;Katayama et al., 2021). Some B. subtilis natto derivatives may be useful to fight against E. faecalis. For example, we have shown in our previous studies that B. subtilis natto can secrete nattokinase mainly to degrade the peptide pheromone cCF10, thereby interfering with the transfer frequency of the antibiotic resistance plasmid pCF10 between E. faecalis bacteria (Lin et al., 2021). In this study, we further demonstrated the effects of B. subtilis natto supernatant on the biofilm formation ability of E. faecalis. In addition, we attempted to clarify the mechanisms by which B. subtilis natto supernatant affects the biofilm formation of E. faecalis using transcriptome sequencing (RNA-seq).

Bacterial Strains, Medium, and Growth Conditions
All bacterial strains used in this study are listed in Table 1. Wild-type E. faecalis OG1RF and constructed E. faecalis OG1RF::p23cfp that expresses a constitutive CFP were obtained from the laboratory of Professor Gary M. Dunny (University of Minnesota, United States). All E. faecalis strains were statically cultured at 37°C in M9 medium [3 g/L yeast extract, 10 g/L casamino acids, 36 g/L glucose, 0.12 g/L MgSO 4 , and 0.011 g/L CaCl 2 (Bandyopadhyay et al., 2016)] or in brain heart infusion broth (BD Co., United States). If needed, the antibiotic rifampicin was added at a concentration of 200 μg/ml (Bandyopadhyay et al., 2016).

Preparation of B. subtilis natto Cell-Free Supernatant
Bacillus subtilis natto supernatant was prepared using methods described in our previous study (Lin et al., 2021). In brief, overnight cultures of B. subtilis natto were diluted 1:100 in M9B medium (8.5 g/L Na 2 HPO 4 ·2H 2 O, 3 g/L KH 2 PO 4 , 11.5 g/L sodium acetate, and 1 ml/L acetic acid were added as buffering agents to the original M9 medium) and cultured aerobically in orbital shakers at 37°C with shaking at 125 rpm. After 24 h of cultivation, the cultures were centrifuged (4,000 × g, 10 min) to remove all cells. Then, the supernatant was filter-sterilized through 0.22-μm filters (Pall Co., United States) and stored at 4°C.

Cell Culture
Caco-2, a colon adenocarcinoma cell line, was purchased from the Bioresource Collection and Research Center (BCRC, Taiwan). The cells were routinely maintained in Dulbecco's modified Eagle medium (DMEM containing 4.5 g/L glucose; GeneDireX Inc., Taiwan) supplemented with 10% fetal bovine serum (FBS; Gibco, United States) and 1x penicillinstreptomycin-glutamine (GeneDireX Inc., Taiwan) and incubated at 37°C in a 95% humidity and 5% CO 2 air atmosphere. The cells were passaged every 5-6 days in 10 cm 2 cell culture dishes.

In vitro Assay of Bacterial Adhesion Onto the Human Intestinal Caco-2 Cell Line
An in vitro bacterial adhesion assay was performed as described by Letourneau et al. with some modifications (Letourneau et al., 2011). To prepare monolayers of Caco-2 cells for the in vitro bacterial adhesion assay, one milliliter of cell suspension (2 × 10 5 cells/ml) was seeded in three sets of duplicate wells (one for each treatment) of a 24-well plate, and the plate was incubated in a cell culture incubator until the cells were fully confluent. The cells were then washed with phosphate-buffered saline (PBS), and the culture medium was replaced with 900 μl of antibiotic-free DMEM supplemented with 10% FBS; then, 100 μl of B. subtilis natto supernatant or M9B medium was added.
Overnight cultures of E. faecalis cells were centrifuged, washed twice with PBS containing 2 mm EDTA, and resuspended in DMEM with 10% FBS. A volume of bacterial culture corresponding to 10 6 E. faecalis cells was used to inoculate Caco-2 cells. The same volume of E. faecalis culture was also added to a medium mixture (90% DMEM containing 10% FBS; 10% M9B medium) without Caco-2 cells to determine the total number of bacterial cells in the inoculum. The E. faecalis and Caco-2 cells were then cocultured at 37°C with 5% CO 2 for 3 h. After 3 h of incubation, the culture medium was removed, and the infected Caco-2 cells were washed 3 times with PBS. All cells were then detached with 0.05% trypsin-EDTA for 20 min at 37°C. Then, serial dilutions of these samples were plated on selective Todd Hewitt broth agar medium containing 30 g/L Bacto Todd Hewitt Broth (Neogen Cor., United States), 15 g/L agar and 50 μg/ml rifampicin, and the adherent E. faecalis cells were counted.

Autoaggregation Assay
An autoaggregation assay was performed as described by Baccouri et al. and Kaur et al. with some modifications (Kaur et al., 2018;Baccouri et al., 2019). Overnight cultures of E. faecalis OG1RF were diluted 1:100 in M9B medium with or without B. subtilis natto supernatant treatment (0, 10, and 50% v/v) and grown in culture tubes with screw caps and rubber liners (Kimble Inc., United States). All culture tubes were incubated anaerobically at 37°C for 24 h. After 24 h of cultivation, treated and untreated E. faecalis cultures were centrifuged, washed twice with PBS, and resuspended in PBS to give final OD 600 of 1. The E. faecalis suspensions were vortexed for 15 s and then incubated at 37°C for 4 h. After 0 and 4 h of incubation without mixing, one milliliter of the suspension from the top of the tube was taken to measure the absorbance (A) at 600 nm. Autoaggregation was then calculated as follows: autoaggregation (%) = [1−(A 4h /A 0 ) × 100].

Experimental Setup for SEM
Overnight cultures of E. faecalis OG1RF were diluted 1:100 in a 24-well plate containing 1 ml of medium and sterile cover glasses and anaerobically cultured at 37°C for 24 or 48 h. The biomass that adhered to the cover glass was prefixed with 2.5% glutaraldehyde and 2.5% paraformaldehyde in 0.05 M cacodylate at 25°C for 60 min.
After prefixation, the samples were washed three times with 0.05 M cacodylate and then postfixed with 1% osmium tetroxide in 0.05 M cacodylate at 25°C for 60 min. The fixed samples were chemically dehydrated using a graded ethanol series [30,50,70,85,90,95, and 100% (two times)] and processed in a CO 2 -based critical point dryer and ion coater. The dried samples were observed using an FEI Inspect S scanning electron microscope at a high voltage of 15 kV and a magnification of 3,000 × .

Biofilm Growth
A ninety-six-well polystyrene plate-based bacterial biofilm formation assay was performed as described by Dale et al. with some modifications (Dale et al., 2015).

Experimental Setup for Confocal Microscopy
Overnight cultures of E. faecalis OG1RF::p23cfp were diluted 1:100 in M9B medium with or without B. subtilis natto supernatant treatment (0, 10, and 50% v/v) and cultured on glass coverslips in 35 × 12 mm tissue culture dishes (Alpha Plus Scientific Co., Taiwan). All dishes were incubated anaerobically at 37°C for 24 or 48 h. After 24 or 48 h of cultivation, the culture medium in the tissue culture dishes was removed, and the biomass attached to the glass coverslips in the tissue culture dishes was washed twice with PBS to remove unattached cells and then fixed with 2% paraformaldehyde (PFA). Fixation with 2% PFA in PBS was performed at 4°C for 10 min. After fixation, the biomass attached to the glass coverslip in the bottom of each petri dish was visualized using a white light laser confocal microscope Leica TCS SP8 X (Leica Microsystems, Ltd., Germany) and analyzed using Leica Application Suite X software.

Extraction and Analysis of Exopolysaccharides From Biofilms
The extraction and analysis of E. faecalis biofilm polysaccharides were conducted using the method described by Liu et al. (2020aLiu et al. ( , 2020b with some modifications. Overnight cultures of E. faecalis were diluted 1:100 in M9B broth with or without B. subtilis natto supernatant (0, 10%, or 50% v/v). One milliliter of diluted E. faecalis culture was added to 24-well plates and incubated anaerobically at 37°C. After 24 or 48 h of cultivation, the culture supernatant was removed, and the biomass that adhered onto the bottom of each well was washed with ddH 2 O and then air-dried for 1 h. Next, the adhered biomass was harvested by scraping the surface thoroughly with a sterile polyester-tipped swab after 1 ml of PBS was added into each well. The cell suspensions of two wells corresponding to the same treatment were mixed together and centrifuged at 5,000 × g for 30 min at 4°C. The concentrated precipitates were resuspended in aqueous solution (2 ml) containing 0.85% NaCl and 0.22% formaldehyde, and the E. faecalis biofilm polysaccharide was extracted at 80°C for 30 min. The polysaccharide dissolved in the formaldehyde solution was recovered further via centrifugation at 15,000 × g and 4°C for 30 min. The polysaccharide concentrations were quantified using the phenol-sulfuric acid (PSA) method (Dubois et al., 1951). In brief, 100 μl of polysaccharide solutions or standard (D-glucose solution) was mixed equally with 5% (w/w) phenol solution in microcentrifuge tubes. Immediately afterward, 1 ml of concentrated sulfuric acid was added. The tubes were then incubated for 5 min at room temperature, and 200 μl of the reaction mixture was added to a 96-well plate. The absorbance was measured at 492 nm using a Multiskan FC microplate photometer (Thermo Fisher Scientific Inc., United States).

HPAEC Analysis of the Carbohydrate Composition of Polysaccharides Obtained From the E. faecalis Cell Envelope
Enterococcus faecalis cell envelope polysaccharides were prepared using the method described by Dale et al. (2015Dale et al. ( , 2017 with some modifications. In brief, overnight cultures of E. faecalis were diluted 1:100 in M9B broth without and with B. subtilis natto supernatant (0 and 10% v/v) and cultured anaerobically.
After 24 h of incubation, E. faecalis cells were collected by centrifugation, and pelleted cells were washed using sucrose solution [25% sucrose and 10 mm Tris-HCl (pH 8)]. Cells were then resuspended in sucrose solution supplemented with lysozyme and mutanolysin and incubated overnight at 37°C with gentle agitation. Next, supernatant fractions were harvested by centrifugation, followed by treatment with RNase A, DNase to remove contaminating nucleic acids, and proteinase K to remove protein impurities. The remaining impurities in the supernatant fraction were further extracted with chloroform. The aqueous phase was transferred to a new tube, and polysaccharides were precipitated by the addition of ethanol to a final concentration of 75% and incubation at −80°C for 30 min. Precipitated polysaccharides were washed with 75% ethanol and allowed to air dry. The carbohydrate composition of E. faecalis cell envelope polysaccharides was analyzed following complete acid hydrolysis of the polysaccharides. Acid hydrolysis of E. faecalis cell envelope polysaccharides was carried out with 1.95 N trifluoroacetic acid (TFA) at 80°C for 6 h. The mixture was cooled, evaporated, and then resuspended in Milli-Q water. Monosaccharides were analyzed on an HPAEC system (Dionex BioLC) equipped with a gradient pump, a pulsed amperometric detector (PAD) using a gold working electrode, and an anion -exchange column (Carbopac PA −10, 4.6 × 250 mm). The detection condition was an isocratic NaOH concentration of 18 mm at ambient temperature. The flow rate was 1.0 ml/min. Identification and quantification of the monosaccharides were carried out in comparison with established standards. Data were collected and integrated on a PRIME DAK system (HPLC Technology, Ltd., United Kingdom).

RNA Purification and Sequencing
Overnight culture of E. faecalis OG1RF was diluted 1:100 in M9B medium with or without B. subtilis natto supernatant treatment and incubated anaerobically at 37°C for 24 h. After 24 h of cultivation, 600 μl of bacterial culture was treated with 1,200 μl of RNAprotect Bacteria Reagent (Qiagen Ltd., Germany) at room temperature for 5 min. The cells were then collected by centrifugation for 10 min at 4°C, flash-frozen in liquid nitrogen and stored at −80°C until RNA extraction. For RNA extraction, cells were lysed with lysozyme (30 mg/ml) and mutanolysin (500 U/ml) in Tris-EDTA (TE) buffer at 37°C for 10 min (Bandyopadhyay et al., 2016;Manias and Dunny, 2018). Total RNA was then extracted using an RNeasy Mini Kit (Qiagen Ltd., Germany) according to the manufacturer's instructions. Five micrograms of total RNA was subjected to DNase treatment with Turbo DNase (Ambion Co., United States) according to the manufacturer's instructions. The RNA purity and concentration were measured using an ND-1000 spectrophotometer (NanoDrop Technologies, Inc., United States), and RNA integrity was validated using a Bioanalyzer 2,100 (Agilent Technologies, Inc., United States). Then, ribosomal RNA was removed using a RiboMinus™ Transcriptome Isolation Kit (Invitrogen Co., United States). The cDNA library was constructed with purified mRNA with a SureSelect Strand Specific RNA Library Preparation Kit (Agilent Technologies, Inc., United States) according to the manufacturer's instructions. RNA-seq using the Illumina NovaSeq 6,000 (paired end) platform was performed at Welgene Biotech (Taiwan).

RNA-seq Data Analysis
The raw image data were converted to sequence data using bcl2fastq conversion software v2.20. After adaptor clipping and sequence quality trimming with Trimmomatic v0.36, the clean reads of the control and treated groups were mapped to the reference genome of the WT strain E. faecalis OG1RF (NC_017316.1) using HISAT2.
Differential expression analysis was performed using cuffdiff (cufflinks v2.2.1) with genome bias detection/correction and Welgene Biotech's in-house pipeline. Genes with a value of p ≤ 0.05 and an FC ≥ 2 were considered significantly differentially expressed. Functional enrichment assays of the differentially expressed genes (DEGs) in each experiment were performed using clusterProfiler v3.6.

RT-qPCR
Total RNA was prepared according to the methods above. Approximately 500 ng of total RNA was used to synthesize cDNA with a Quant II Fast RT kit (BIOTOOLS Co., Ltd., Taiwan). Each reaction product was diluted 1:20 in ddH 2 O, and 1 μl was then used for quantitative real-time polymerase chain reaction (RT-qPCR) using SYBR ® Green Supermix (Bio-Rad Laboratories, Inc., United States) and a Bio-Rad CFX384 instrument (Bio-Rad Laboratories, Inc., United States). A total reaction volume of 10 μl containing 2 μl of gene-specific primer mixture at a concentration of 2 μm was used in each well. Each reaction was performed in triplicate, and the threshold cycle (C T ) values were obtained. Relative quantification was performed using the 2 -ΔΔCT method (Livak and Schmittgen, 2001). Gene expression was normalized to that of the housekeeping gene gyrB, which is moderately expressed (Chatterjee et al., 2013). The sequences of the primers are listed in Table 2. Three biological replicates were performed to show repeatability.

Bacillus subtilis natto Supernatant Affects E. faecalis Adhesion Onto Human Intestinal Caco-2 Cell Monolayers
Bacterial adhesion to and subsequent colonization of host tissues are key steps in the initial stage of biofilm formation (Baldassarri et al., 2001;Monte et al., 2014). To test our hypothesis that B. subtilis natto supernatant would interfere with E. faecalis biofilm formation, we first investigated whether the ability of E. faecalis to adhere onto the human intestinal epithelial cell line Caco  (Figures 1B,D). These results indicated that B. subtilis natto supernatant likely interferes with the ability of E. faecalis to adhere to host tissues, such as the human intestinal tract.

Bacillus subtilis natto Supernatant Reduces the Formation of E. faecalis Aggregates
Previous studies (Kragh et al., 2016) have shown that bacterial cells tend to clump together in multicellular aggregates before they form biofilms. Herein, we investigated whether B. subtilis natto supernatant affects the formation of E. faecalis aggregates. Overnight cultures of E. faecalis were diluted in M9B medium in the presence or absence of the B. subtilis natto supernatant and incubated under anaerobic conditions. After 24 h of cultivation without mixing, both 10 and 50% (v/v) B. subtilis natto supernatant-treated E. faecalis cultures remained turbid, whereas control (untreated) cultures settled at the bottom of the tube (Figure 2A). Treated and untreated E. faecalis cells were collected and used in 4-h autoaggregation assay. As shown in Figure 2B, percentage autoaggregation of both 10 and 50% (v/v) B. subtilis natto supernatant-treated E. faecalis was significantly lower than that of untreated E. faecalis (control group; 10%: p < 0.05; 50%: p < 0.01). In addition, the aggregates of E. faecalis cells formed on glass coverslips with or without B. subtilis natto supernatant were observed and analyzed using scanning electron microscopy (SEM). As shown in Figure 2C, E. faecalis cells in the control group formed dense aggregates after 24 h of cultivation. However, the dense aggregation of E. faecalis cells was markedly reduced upon treatment with 10 and 50% v/v B. subtilis natto supernatant. Even though E. faecalis grew for 48 h, no large cell clumps were observed in the B. subtilis natto supernatant-treated groups. Collectively, these results indicated that B. subtilis natto supernatant might inhibit E. faecalis autoaggregation, thus affecting biofilm initiation and development.

Inhibition of E. faecalis Biofilm Production by B. subtilis natto Supernatant
Ninety-six-well polystyrene plate-based biofilm assays devised by Dale et al. (2015) were performed here to investigate the effect of B. subtilis natto supernatant on E. faecalis biofilm growth over 24 h. Overnight cultures of E. faecalis were diluted in M9B medium in the presence or absence of the B. subtilis natto supernatant (10% v/v) and incubated under anaerobic conditions for 24 h. The optical density at 600 nm (OD 600 ) values of the E. faecalis cultures were measured  Figure 3C). These results showed that B. subtilis natto supernatant could inhibit E. faecalis biofilm production without killing planktonic E. faecalis cells.
The difference between the B. subtilis natto supernatanttreated and untreated groups seemed to be more significant in SEM images ( Figure 2C) than in the biofilm formation assay (Figure 3C). In SEM experiments, E. faecalis cells attached to glass coverslips that had no coating, whereas E. faecalis cells in the biofilm formation assay attached to the tissue culture-treated bottom of a 96 well polystyrene plate. We found that E. faecalis cells had relatively difficulty attaching to noncoating abiotic surfaces, such as glass coverslips. Furthermore, some E. faecalis cells or clumps that attached to glass coverslips may be removed due to multistep sample pretreatment before SEM observations. Therefore, we inferred that these factors might contribute to the deviation between the results of the SEM experiment , the data are presented as the means ± SDs (n = 3). *p < 0.05 and **p < 0.01 compared with each negative control group (Student's t-test). and the 96-well plate-based biofilm formation assay (Figures 2C, 3C, respectively).

Enterococcus faecalis Biofilm Architecture Is Impacted by B. subtilis natto Supernatant
The 3D architecture of the biofilms of the CFP-expressing strain E. faecalis OG1RF::p23cfp that formed on the glass coverslips at the bottom of the tissue culture dishes with or without B. subtilis natto supernatant was visualized and analyzed using confocal laser scanning microscopy (CLSM). The results are shown in E. faecalis were thinner (at almost 12 and 5 μm, respectively) than the E. faecalis biofilm in the control group, which was 18 μm thick. Although all the control and B. subtilis natto supernatant-treated biofilms grew thicker from 24 to 48 h, the inhibitory effect of B. subtilis natto supernatant on E. faecalis biofilms still existed (approximate biofilm thickness: 25 μm in the control group, 18 μm in both treated groups).
The CLSM results were consistent with the results shown by biofilm assays conducted in 96-well polystyrene plates. Taken together, these results indicated that B. subtilis natto supernatant interfered with E. faecalis biofilm growth and resulted in the formation of looser and thinner biofilms.

Inhibition of E. faecalis Biofilm Polysaccharide Production by B. subtilis natto Supernatant
We found that E. faecalis cells can easily and effectively attach to the bottom of tissue culture-treated plates or dishes and form biofilms. Therefore, in the 96-well plate-based biofilm formation assay, we hypothesized that approximately the same number of E. faecalis cells attached to the tissue culture-treated surface in the presence or absence of B. subtilis natto supernatant. The results in Figure 3C show that E. faecalis biofilm production was inhibited by B. subtilis natto supernatant treatment. In addition, the E. faecalis Relative biofilm production (%) was calculated by further normalizing the biofilm index values of the treated group to those of the negative control group to which no B. subtilis natto supernatant was added. The data are presented as the means ± SDs (n = 8). **p < 0.01 and ***p < 0.001 compared with each negative control group (Student's t-test).
biofilms in the B. subtilis natto supernatant-treated groups were thinner than those in the control (untreated) groups under CLSM (Figure 4). However, B. subtilis natto supernatant did not inhibit the growth of E. faecalis (Figures 3A,B). Based on these observations, we inferred that the secretion of extracellular polymeric substances (EPSs) in E. faecalis biofilms might be changed. When bacteria adhere to a solid surface, they continue to grow and secrete EPS, which comprises polysaccharides, proteins, nucleic acids, and fatty acids (Liu et al., 2020a,b). EPS plays a key role in forming the three-dimensional architecture of biofilms. As reported previously, EPS in most biofilms accounts for more than 90% of the dry mass (Flemming and Wingender, 2010). Furthermore, polysaccharides are the major components of EPS (Flemming and Wingender, 2010). Herein, we investigated whether the polysaccharide contents of E. faecalis biofilms were reduced. The quantification results in Figure 5 show that the polysaccharide contents of E. faecalis biofilms indeed decreased in both the 10 and 50% (v/v) B. subtilis natto supernatant-treated groups after 24 h of cultivation. When E. faecalis grew for 24 to 48 h, the biofilm exopolysaccharide contents in all the control and treated groups increased. However, the polysaccharide contents in both the 10 and 50% (v/v) B. subtilis natto supernatant-treated groups were significantly lower than those in the control groups. Thus, B. subtilis natto supernatant likely inhibited E. faecalis biofilm polysaccharide production.

Bacillus subtilis natto Supernatant Restructures Carbohydrates in the E. faecalis Cell Envelope
Previous studies (Haussler et al., 2003;Kragh et al., 2016) have reported that greater cell surface "stickiness" may increase the tendency of bacterial cells to form aggregates and attach to a solid surface, thus increasing biofilm formation. The properties of the bacterial cell surface, including stickiness, may be associated with the composition and organization of the bacterial cell envelope, which comprises the inner cell membrane and the cell wall (Sengupta et al., 2013;Choi et al., 2015). Here, we investigated whether the carbohydrate composition of the E. faecalis cell envelope was impacted by B. subtilis natto supernatant treatment (10% v/v, 24 h) using high-performance anion-exchange chromatography (HPAEC). As shown in Figure 6, several types of monosaccharides, including glucose, glucosamine, and galactosamine, were detected in the E. faecalis cell envelope. Glucose, which accounted for approximately 45 mol % of all monosaccharides in the E. faecalis cell envelope, was the most abundant monosaccharide. Upon treatment with B. subtilis natto FIGURE 4 | 3D architecture of E. faecalis OG1RF::p23cfp biofilms with or without B. subtilis natto supernatant. The CFP-expressing strain E. faecalis OG1RF::p23cfp was cultured anaerobically in M9B medium without or with B. subtilis natto supernatant (0, 10, or 50% v/v) on glass coverslips in tissue culture dishes. After 24 or 48 h of cultivation, the biofilms that formed on the bottom of the culture dishes were washed, fixed, and then analyzed using CLSM. The approximate biofilm thicknesses (μm) for all groups were measured and are shown in the figure.
FIGURE 5 | Inhibitory effect of B. subtilis natto supernatant on E. faecalis biofilm polysaccharide production. E. faecalis OG1RF was anaerobically cultured in M9B broth without or with B. subtilis natto supernatant (0, 10, and 50% v/v) for 24 or 48 h. The polysaccharide contents obtained from E. faecalis biofilms were determined using the PSA method. The data are presented as the means ± SDs (n = 3). Values with asterisks (*) were significantly different compared with each negative control group according to Duncan's multiple range tests (*p < 0.05).

RNA-seq Analysis of E. faecalis in Response to B. subtilis natto Supernatant
To further understand how E. faecalis responded to B. subtilis natto supernatant, we searched for DEGs between E. faecalis treated with or without B. subtilis natto supernatant using RNA-seq analysis. According to the RNA-seq results, approximately 95.94 and 97.98% of the clean reads from the treated and control (untreated) groups were mapped to the reference genome, respectively. Among the 2,657 genes detected by RNA-seq [2,658 genes present in E. faecalis OG1RF (Bourgogne et al., 2008)], 95 genes were identified as differentially expressed [value of p ≤ 0.05 and fold change (FC) ≥ 2] in the treated group (Figure 7 and Supplementary Table 1). Among these DEGs, 70 genes were identified as significantly downregulated, and 25 genes were found to be significantly upregulated. Furthermore, Gene Ontology (GO) annotation analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed to identify DEGs at the biological function level. The GO annotation results showed that the genes associated with the ATP binding term were differentially expressed, and the KEGG pathway mapping results showed that some DEGs were involved in the peptidoglycan biosynthetic process (Supplementary Figures 1, 2). These results indicated that some genes encoding ATP binding protein or peptidoglycan biosynthetic genes might be differentially expressed due to B. subtilis natto supernatant treatment.
To verify the RNA-seq results, we selected a total of 13 DEGs, including some ATP binding protein-encoding genes and peptidoglycan biosynthetic genes, for further RT-qPCR analysis. As listed in Table 3, the up-or downregulation trends and relative expression of most selected DEGs, including three peptidoglycan biosynthetic genes [murD (OG1RF_ RS03760), murN (OG1RF_RS08680), and uppS (OG1RF_ RS09795)], bgsA (OG1RF_RS11220), and walK (OG1RF_ RS05035), were consistent with the RNA-seq results. Furthermore, a correlation between the six downregulated DEGs listed above and E. faecalis cell wall homeostasis and biofilm formation has been reported previously (Dubrac et al., 2008;Vollmer et al., 2008;Theilacker et al., 2009;Watanabe et al., 2012;Bucher et al., 2015). Peptidoglycan, an essential component of the cell wall in almost all bacteria, plays a key role in maintaining cell shape and serves as a scaffold to anchor other cell envelope components, such as proteins and teichoic acids (Vollmer et al., 2008). Gene bgsA (OG1RF_ RS11220) encodes the putative glucosyltransferase designated  Frontiers in Microbiology | www.frontiersin.org biofilm-associated glycolipid synthesis A and synthesizes diglucosyl-diacylglycerol (DGlcDAG) in E. faecalis (Theilacker et al., 2009). DGlcDAG, a precursor of glycolipid and lipoteichoic acid, is involved in E. faecalis biofilm production, adherence to host cells and virulence in vivo (Theilacker et al., 2009). Gene walK (OG1RF_RS05035) encodes the membrane-linked histidine kinase WalK. WalK is a key element in the WalK/ WalR two-component signal transduction system, which regulates genes involved in cell wall metabolism, biofilm production, virulence regulation, oxidative stress resistance, and antibiotic resistance in low-G + C Gram-positive bacteria, including E. faecalis (Dubrac et al., 2008;Watanabe et al., 2012). Based on these findings, we focused on the six downregulated cell wall-and biofilm-related genes in E. faecalis in further investigations.

Inhibitory Effect of B. subtilis natto Supernatant on the E. faecalis Peptidoglycan Biosynthesis Pathway and WalK/WalR Two-Component System
To investigate whether B. subtilis natto supernatant inhibits the expression of other genes in the peptidoglycan biosynthesis pathway and WalK/WalR regulatory system, the relative expression of UDP-GlcNAc enolpyruvyl transferase (murA; OG1RF_ RS10140), murB (OG1RF_RS10725), murF (OG1RF_RS03015), and murG (OG1RF_RS03765) in the peptidoglycan biosynthesis pathway and the response regulator-encoding gene walR (OG1RF_RS05030) in the WalK/WalR two-component system were assessed further using RT-qPCR. As shown in Figure 8, the relative expression of several peptidoglycan biosynthetic genes except murF was indeed inhibited due to B. subtilis natto supernatant treatment. The relative expression of walR in the WalK/WalR two-component system was also inhibited. Taken together with the RT-qPCR results of the six cell wall-and biofilm-related genes identified from RNA-seq, these results indicate that B. subtilis natto supernatant might interfere with the E. faecalis peptidoglycan biosynthesis pathway, WalK/WalR two-component system and bgsA, thus inhibiting biofilm formation.

DISCUSSION
In this study, we found that the culture supernatant of B. subtilis natto can potently inhibit E. faecalis adherence to Caco-2 cell monolayers, aggregation, and biofilm production. These findings support the idea that some derivatives present in B. subtilis natto supernatant may have antibiofilm activity against E. faecalis. Consistent with this idea, some derivatives of Bacillus species have been shown to inhibit the formation of biofilms by several bacterial pathogens (Rivardo et al., 2009;Yoo et al., 2019;Tazehabadi et al., 2021). Rivardo et al. (2009)  Streptococcus mutans. Tazehabadi et al. (2021) mentioned that the bacteriocins subtilosin A and subtilin produced by B. subtilis and B. amyloliquefaciens, respectively, play a key role in inhibiting Salmonella biofilm formation (Yoo et al., 2019). We also found that E. faecalis produced fewer exopolysaccharides under treatment with the culture supernatant of B. subtilis natto than under control conditions. Previous studies (Ramos et al., 2019) have shown that β-1,6-linked poly-N-acetylglucosamine (polyGlcNAc) in exopolysaccharides produced by E. faecalis bacteria enables the bacteria to successfully penetrate semisolid surfaces and translocate through human epithelial cell monolayers (Ramos et al., 2019). Thus, the inhibitory effect on E. faecalis exopolysaccharide production might also contribute to reducing polyGlcNAc production, thus mitigating the pathogenicity of E. faecalis.
In addition to exerting an inhibitory effect on exopolysaccharide production, B. subtilis natto supernatant restructured carbohydrates in the E. faecalis cell envelope. A previous study reported that the structure and composition of the bacterial cell envelope are linked to bacterial cell surface properties, such as surface stickiness (Sengupta et al., 2013). Another study reported that bacterial cells with greater surface stickiness have a tendency to aggregate (Kragh et al., 2016). In addition, the increased tendency to aggregate has been shown to be associated with increased biofilm production in Pseudomonas aeruginosa (Deziel et al., 2001;Haussler et al., 2003). Based on these findings, we infer that B. subtilis natto supernatant may alter E. faecalis cell envelope composition and cell surface properties, thereby interfering with E. faecalis adhesion to host tissues, aggregation, and biofilm production.
Via RNA-seq and RT-qPCR analysis, we found that B. subtilis natto supernatant inhibited the expression of bgsA, which encodes biofilm-associated glycolipid synthesis A, in E. faecalis. Previous studies (Theilacker et al., 2009) have shown that inactivation of bgsA in E. faecalis leads to a lack of DGlcDAG in cell membranes and to accumulation of longer lipoteichoic acid structures in cell walls, thus impairing E. faecalis adherence to host tissues and biofilm production. In addition to bgsA, walK and walR in the WalK/WalR two-component system were inhibited in E. faecalis treated with B. subtilis natto supernatant. In low-G + C Gram-positive bacteria, the activation of the WalK/WalR two-component system is essential for lateral cell wall synthesis and cell elongation, whereas depletion of this system may cause defects in cell morphology, murein (peptidoglycan) synthesis, and biofilm formation (Dubrac et al., 2008;Takada and Yoshikawa, 2018). Moreover, we also found that several peptidoglycan biosynthetic genes (murA, murB, murD, murG, murN, and uppS) in E. faecalis were inhibited. Since peptidoglycan is a crucial structural element in the cell walls of most bacteria, interference with its biosynthesis results in impaired biofilm formation and even cell lysis (Vollmer et al., 2008;Bucher et al., 2015). The downregulated genes listed above are all related to cell envelope synthesis and biofilm formation in E. faecalis. The evidence suggests that B. subtilis natto supernatant targets E. faecalis cell envelope synthesis and therefore interferes with the cell envelope composition and biofilm formation of E. faecalis.
Our findings reveal that B. subtilis natto supernatant can likely inhibit biofilm formation of E. faecalis via interference with E. faecalis cell envelope synthesis. Notably, the bacterial cell envelope synthetic process has been reported to be the major target for many antibacterial agents (McCallum et al., 2011). Many antibiotics act by blocking or disrupting bacterial peptidoglycan biosynthesis, such as fosfomycin, which inhibits MurA (Hashemian et al., 2019), and tunicamycin, which inhibits bacterial phospho-N-acetylmuramic acid (MurNAc)pentapeptide translocase (MraY; Yamamoto et al., 2019). Furthermore, because maintenance of cell wall homeostasis and growth in low-G + C Gram-positive bacteria is essential, the WalK/WalR two-component system has been proposed as a novel target for antibacterial agents that are effective against multidrug-resistant bacteria, including MRSA and vancomycin-resistant E. faecalis (Fabret and Hoch, 1998;Gotoh et al., 2010). Watanabe et al. (2012) reported that the novel antibiotic signermycin B from Streptomyces extracts can function as a WalK inhibitor, targeting the conserved dimerization domain of WalK to inhibit autophosphorylation (Watanabe et al., 2012). Collectively, these studies provide evidence that some substances that function like antibiotics or WalK inhibitors may be present in the culture supernatant of B. subtilis natto. Further investigations are necessary to identify potential antibiofilm agents in the culture supernatant of B. subtilis natto.
The therapeutic potential of B. subtilis and its derivatives in animals has been reported in previous studies (Cartman et al., 2008;Piewngam et al., 2018). One previous study (Cartman et al., 2008) showed that B. subtilis spores can germinate in the chicken gastrointestinal tract. Another study (Piewngam et al., 2018) showed that mice fed B. subtilis spores exhibit complete decolonization of MRSA in the feces and intestines. These studies provide evidence that B. subtilis natto spores may germinate to form vegetative cells and produce functional substances in host gastrointestinal tracts.
In this work, our results showed that B. subtilis natto derivatives present in the culture supernatant could effectively inhibit the formation of E. faecalis biofilms. These derivatives downregulated the transcription of genes involved in membrane glycolipid biosynthesis, the WalK/WalR two-component system, and peptidoglycan biosynthesis, which may contribute to changes in the structural components of the cell envelope and therefore affect biofilm formation ability in E. faecalis. Based on these findings, we propose that natto or the probiotic B. subtilis natto could be used in the management of E. faecalis biofilm infections.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found online at: https://www.ncbi. nlm.nih.gov/geo/, GSE184249.

AUTHOR CONTRIBUTIONS
Y-CL wrote the manuscript. K-TL, W-SH, and Y-CL designed the experimental plan. Y-CL and C-YW performed all experiments and analyzed the relevant data. H-TH and M-KL assisted with carbohydrate composition analysis using HPAEC and data interpretation. All authors contributed to the revision and final review of the manuscript.