Cell wall associated protein TasA provides an initial binding component to extracellular polysaccharides in dual-species biofilm

Many bacteria in biofilm surround themselves by an extracellular matrix composed mainly of extracellular polysaccharide (EP), proteins such as amyloid-like fibers (ALF) and nucleic acids. While the importance of EP in attachment and acceleration of biofilm by a number of different bacterial species is well established, the contribution of ALF to attachment in multispecies biofilm remains unknown. The study presented here aimed to investigate the role of TasA, a precursor for ALF, in cell-cell interactions in dual-species biofilms of Bacillus subtilis and Streptococcus mutans. Expression of major B. subtilis matrix operons was significantly up-regulated in the presence of S. mutans during different stages of biofilm formation, suggesting that the two species interacted and modulated gene expression in each other. Wild-type B. subtilis expressing TasA adhered strongly to S. mutans biofilm, while a TasA-deficient mutant was less adhesive and consequently less abundant in the dual-species biofilm. Dextran, a biofilm polysaccharide, induced aggregation of B. subtilis and stimulated adhesion to S. mutans biofilms. This effect was only observed in the wild-type strain, suggesting that interactions between TasA and dextran-associated EP plays an important role in inter-species interactions during initial stages of multispecies biofilm development.


B. subtilis and S. mutans form complex structures within dual-species biofilm.
The first aim of this study was to determine whether an interaction between B. subtilis and S. mutans cells may result in formation of dual-species biofilm. We compared the morphology of mono and dual-species biofilms using scanning electron microscopy (SEM). S. mutans mono-species biofilm was homogenous and surrounded by EP. B. subtilis mono-species biofilm contained bundles of filamentous cells. The dual-species biofilm had more three-dimensionally complex structures than the mono-species biofilm (Fig. 1, right panel). Moreover, our observations showed close proximity between the rod shaped B. subtilis and the S. mutans cocci (Fig. 1).

B. subtilis cell initially interacts with S. mutans biofilm via TasA. The co-localization of B. subtilis
and S. mutans in the dual-species biofilm indicated a direct interaction between these bacterial cells. To explore this interaction, we quantified adhesion forces between the two species by single cell force spectroscopy (SCFS) using an atomic force microscope (AFM). We quantified adhesion forces between single B. subtilis cells and a glass coverslip surface (control sample), a B. subtilis biofilm and a S. mutans biofilm.
The first step was the attachment of a single B. subtilis cell onto a tipless cantilever coated with polydopamine (PDA) (Supplementary Fig. S1). Next, the cantilever was repeatedly approached to and retracted from the test surface to determine the adhesion energy, the maximum adhesion force, and the distance at which the bond ruptures as the cantilever is retracted from the surface (Fig. 2 a1-4). While the force profile of B. subtilis on glass featured weak interactions at short distances, the force profile between a single B. subtilis cell and B. subtilis biofilm showed multiple rupture events at longer distances. The force profiles of B. subtilis cells and S. mutans biofilm showed increasing number of rupture events, higher adhesion force, and longer rupture length, suggesting interaction with multiple adhesins that could be stretched from the cell surface. The strong interaction between B. subtilis and S. mutans was also reflected in the number of adhesion events observed in force spectroscopy analysis. B. subtilis adhered to S. mutans biofilm in 100% of the measured force curves, while up to 40% of force curves showed no adhesion when B. subtilis was approached to a glass surface or a B. subtilis biofilm.
Adhesion is described by the maximum adhesion force measured between the bacterium and the sample ( Fig. 2b-1), and the adhesion energy calculated from the area under the force-distance curve ( Fig. 2b-2). The adhesion force between the B. subtilis cell and the B. subtilis biofilm was 2-fold higher than the adhesion force to glass. However, its adhesion force to S. mutans biofilm was 5-fold higher (Fig. 2b). The adhesion force between B. subtilis cells and S. mutans biofilm was so high that it frequently caused detachment of B. subtilis from the cantilever.
It has been established that TasA is a cell wall associated protein that plays a crucial role in B. subtilis' biofilm formation 20 . Therefore, the next step was to examine the possible role of TasA in interaction between B. subtilis and other bacteria in mixed-species biofilm. Using SCFS, we were able to measure the initial adhesion force of ΔtasA mutant strain of B. subtilis to S. mutans biofilm. The force profiles obtained for the ΔtasA mutant towards glass (data nor shown) or S. mutans biofilm were similar to those of the WT strain ( Fig. 2a-4). However, in contrast to the WT B. subtilis, the ΔtasA mutant showed weak interaction with S. mutans biofilm (Fig. 2b).
TasA affects the structure and the incorporation of B. subtilis into co-species biofilm. Next, confocal laser scanning microscopy (CLSM) was used to observe the three-dimensional structure of dual-species biofilm. In Fig. 3, dextran-associated EP, formed by S. mutans, is labelled with dextran alexa fluor (blue), and B. subtilis cells are tagged with GFP (green). The CLSM images indicate that B. subtilis cells were not only attached to the upper layer of the biofilm but also appear in deeper layers and are surrounded by dextran-associated EP SCIeNTIFIC REPORTS | (2018) 8:9350 | DOI:10.1038/s41598-018-27548-1 (Fig. 3a). However, B. subtilis ΔtasA cells (green) were less abundant in the dual-species biofilm (Fig. 3b) after 24 hours. B. subtilis (both in WT and ΔtasA) increased in abundance after 48 hours, but ΔtasA remained less abundant than the WT B. subtilis (Fig. 3). These results indicate that TasA contributes to establishment of B. subtilis in co-species biofilm with S. mutans, especially in the initial stages.
TasA is essential for dextran-induced B. subtilis aggregation. S. mutans' EP serve as a platform for bacterial adhesion 25 . Therefore, we tested whether dextran (a primary component in S. mutans' EP 23,26 ) contributes to the attachment of B. subtilis cells. An aggregation assay showed that addition of increasing concentrations of dextran stimulated aggregate formation by WT cells of B. subtilis by 40-80%, while dextran supplementation could not notably affect the ΔtasA mutant cells (Fig. 4). The strains had similar aggregation ratio in the absence of dextran (data not shown).

The amount of dextran-associated EP affects incorporation of B. subtilis cells to the dual-species biofilm.
To investigate whether dextran affects B. subtilis biofilm formation, we determined if enzymatic removal of dextran during biofilm formation affected the abundance of B. subtilis. We visualised B. subtilis WT cells in dual-species biofilm formed in the presence of different dextranase concentrations. The abundance of B. subtilis biomass decreased in a dose-dependent manner in the dual-species biofilm in response to increasing concentration of dextranase (Fig. 5), suggesting that the quantity of dextran is important for B. subtilis adherence in dual-species biofilm.
Expression of tapA operon of B. subtilis is up-regulated in the presence of S. mutans. Finally, we aimed to determine whether the presence of S. mutans affects expression of the tapA-sipW-tasA operon. The β galactosidase assay showed approximately 5-fold up-regulation in the activity of the tapA promoter during the first 12 h of co-culture, although the activity decreased over time. S. mutans is a very acidogenic bacterium which produces a high amount of acid during its growth. We therefore tested, if its effect on gene expression was through modifying the pH in the media. A shift in the pH from 7 to 6 up-regulated the tapA promoter activity, although to a lesser extent than the up-regulation observed in the presence of S. mutans (Fig. 6a).
Since ΔtasA mutant appeared in the dual-species biofilm after 24 hours and furthermore, the increase in the amount of the mutant cells after 48 hours indicated that another factor may be involved in the attachment of B. subtilis to the dual-species biofilm. Therefore, the next step was to determine whether the presence of S. mutans affects the expression of the epsA-O operon. The β galactosidase assay showed approximately 2.5-fold up-regulation in the expression of the eps promoter during the first 12 h of co-culture. In contrast to the tapA operon, the highest up-regulation in the eps operon was only after 24 hours. These results indicate that the regulation of both of operons is not stimulated simultaneously in the dual-species biofilm.

Discussion
The current study aimed to explore the role of TasA in the interaction between B. subtilis and S. mutans in dual-species biofilm. Our results point to an important role for interaction between two matrix components -B. subtilis' TasA protein and S. mutans' dextran polysaccharides -during dual-species biofilm formation.
Biofilm formation begins with attachment of bacteria to the substrate. Attachment involves initial reversible adhesion, which occurs within seconds to minutes, followed by irreversible adhesion, which results in bacterial aggregation 27,28 . The high sensitivity and resolution of the AFM allows quantification of adhesion forces of a single bacterial cells to probe the interaction forces that govern initial attachment. Therefore, the initial reversible interaction of B. subtilis and S. mutans cells was determined using SCFS. The low adhesion force between B. subtilis and glass explains the inability of B. subtilis NCIB3610 to adhere to surfaces and to form submerged biofilm at solid-liquid interfaces 29 . B. subtilis adhered more strongly to B. subtilis biofilm, but its adhesion force to S. mutans biofilm was significantly stronger. The relatively high inter-species adhesion force explains the capability of B. subtilis to form submerged biofilm in cooperation with S. mutans, which more easily colonizes the glass substrate.
The importance of TasA in biofilm formation by B. subtilis is well established in the literature 20,22 . Additionally, the role of ALF in cell-host adhesion was demonstrated in other bacterial species such as Escherichia coli and Salmonella enteritidis 30,31 . We show here that TasA was critical for the interaction between B. subtilis cells to S. mutans biofilm, as a TasA deficient mutant adhered only weakly to S. mutans (Fig. 2) and was less abundant in the dual-species biofilms (Fig. 3).
The role of TasA in the initial formation of dual-species biofilm raised the question of whether the presence of S. mutans affected the expression of tasA. Expression of the tapA-sipW-tasA operon increased in the presence of S. mutans during the first 12 h of co-culture, and subsequently decreased after 36 h, implying that TasA is required in early stages of dual-species biofilm development. Bacteria are known to differentiate into biofilm-forming cells in response to environmental cues 32,33 . Thus, induction in the expression of tapA-sipW-tasA operon may be triggered by the production of an organic acid(s) by S. mutans and a decrease in pH levels due to sucrose metabolism 34,35 . Therefore, tapA-sipW-tasA expression was examined in a medium with decreased pH by addition of lactic acid, a major end-product of glycolysis by S. mutans 36 . The tapA expression was up-regulated at pH 6, indicating a pH-dependent induction of the tapA operon. These results support the findings by Chen et al. 37 , who found that organic acids stimulate biofilm formation by B. subtilis 37 .   suggesting that the interaction between TasA and S. mutans biofilm is due to ALF-EP interaction. This result is supported by Romero et al. 22 , who suggest that TasA ALF interact in a specific way with EP matrix 22 and Besingi et al. 40 who suggested that amyloid proteins can be adhesive to glucans 40 . The aggregation of B. subtilis in the presence of dextran demonstrates the irreversible interactions that occur between the bacteria. Our results support the finding of Diehl et al. 41 who suggested, based on nuclear magnetic resonance (NMR) spectra, that TasA interacts specifically with EP. Such interactions can explain the strong adhesion of B. subtilis cells to S. mutans biofilm, as well as the reduced adherence of the ∆tasA strain. Moreover, their study showed that EP promotes the restructuring of TasA into ALF, which may explain the increased aggregation of B. subtilis WT in the presence of Dextran 41 .
A recent study by Besingi et al. 40 found amyloid fibers in S. mutans biofilms. However, we did not detect any ALF in the S. mutans biofilms. This could be due to the use of sucrose in the growth media, as glucans formed  B. subtilis (P tapA-lacZ or P eps-lacZ ) cells were grown in the presence or absence of S. mutans cells either in pH7 or in pH6. Samples were collected every 12 hours for β galactosidaze assay. The data are displayed as a mean value ± standard deviation of data from three biological repeats each performed as triplicates. The expression of the tasA and the eps operons was significantly induced in the presence of S. mutans. The expression time of both of the operons was not in parallel. While tasA expression was significantly increased after the first 12 h of incubation, the higher induction in eps expression was after 36 h. *P value < 0.05 compared to B. subtilis pH = 7. from sucrose can mask amyloid production 40 . However, the existence of such fibers may also contribute to the adhesion of more bacterial species and the complexity of the formed biofilm in multispecies settings.
Overall, this study demonstrates that the interaction between TasA produced by B. subtilis and EP produced by S. mutans plays an essential role in initial attachment of bacteria during dual-species biofilm formation. Moreover, we demonstrate the complexity of how each matrix component can influence multispecies biofilm formation. This study provides new insights into ALF contribution to the structural complexity of dual-species biofilm. A clear view of this interaction may result in new approaches to control biofilm formation and adhesion of these specific bacterial species. Moreover, further investigation of the specific active sites on such extracellular matrix components may lead to developing targeted treatments to combat biofilms.

Materials and Methods
Strains and growth media. The cultures of clinical isolate strain S. mutans UA159 were grown overnight in brain heart infusion broth (BHI, Acumedia, Landing, Michigan) at 37 °C in 95% air/5% CO 2 . The following B. subtilis strains were used in this study: NCIB3610 (WT), YC161 (P spank -gfp), DDA (∆tasA, P spank -gfp), RL4582 (P tapA -lacZ) and BS505 (∆tasA). For starter culture generation, one colony of B. subtilis from fresh Lysogeny broth (LB, Neogen, Lansing, Michigan, USA) agar plate was grown in LB and incubated at 37 °C at 150 rpm for 5 h. For all the experiments, both bacteria species were collected in late exponential phase.
Mono and dual-species biofilm formation. For B. subtilis mono-species biofilm, an overnight culture (30 °C at 150 rpm) was diluted 1:10 (to obtain final O.D (600 nm) = 0.07) into LB containing 3% lactose, and incubated at 37 °C at 50 rpm for 5 h 42 . For S. mutans mono-species biofilm, in 96 well plate (Nunc, Roskilde, Denmark), overnight cultured S. mutans (O.D (600 nm) = 1) were diluted 1:10 into fresh BHI supplemented with 2% sucrose. The plate was incubated at 37 °C in 95% air/5% CO 2 for 24 or 48 h. The medium was replaced with fresh medium after 24 h of incubation 34 . For dual-species biofilm, cells of B. subtilis and S. mutans were grown in 96 well plate as follows: to fresh BHI supplemented with 2% sucrose were introduced the overnight cultured S. mutans (diluted by a ratio 1:10, approximately 2.5 × 10 7 CFU) and B. subtilis (diluted by a ratio 1:100, approximately 2.5 × 10 5 CFU). The ratio between B. subtilis and S. mutans cells that were seeded to obtain the dual-species biofilm was approximately 1:100. The plate was incubated at 37 °C in 95% air/5% CO 2 for 24 or 48 h. The medium was replaced with fresh medium after 24 h.

Visualization of mono and dual-species biofilms by SEM. SEM analysis was used for initial visualiza-
tion of the generated biofilms mainly in terms of structure and the spatial organizations in multispecies biofilm. For B. subtilis mono-species biofilm: a sample of 10 µl from the culture containing the B. subtilis mono-species biofilm was placed on small cover slip 43,44 . The S. mutans mono and dual-species biofilms were grown as described above on sterile coverslip (5 mm diameter, glass, Barnaor, Israel). The coverslips were removed and washed twice with sterile demineralized water (DDW). All samples were fixed using 4% formaldehyde for 20 min and washed using DDW. The samples were visualized using an analytical Quanta 200 Environmental High Resolution Scanning Electron Microscope (EHRSEM) (FEI, Eindhoven, The Netherlands).
Single cell force spectroscopy. SCFS was used to determine the initial interaction between B. subtilis cell and S. mutans biofilm. The experiment was conducted as described by Zeng et al. 45 , previously with slight modifications 45 . Biofilm preparation. S. mutans biofilm was prepared as described above. The coverslip was removed, washed twice with PBS and glued to glass-slide using epoxy (JPK, Germany).
AFM force measurements. Force measurements were conducted in PBS using JPK Nanowizard 3 (JPK instruments, Germany). The sensitivity of tipless cantilevers was calibrated on a glass surface in PBS and the spring constant was calibrated using the thermal fluctuation method (included in the AFM software). After immobilization of a bacterial cell on the cantilever, force spectroscopy was performed by approaching the cantilever with the immobilized bacterium to either bare glass or biofilms of B. subtilis/S. mutans, to determine the maximum adhesion force of the cell to each sample. Force curves were measured with a maximum force of 2.5 nN and 15 s extend delay and repeated 10 times at three different positions. Measurements with each bacterium were conducted for a period of approximately 20 min. To ensure that the bacterium was still attached to the cantilever and that it retained its position, optical microscopy was used during the measurements when possible (glass, and B. subtilis biofilm). For measurements on S. mutans biofilm, visualization of the bacteria on the probes was conducted before and after each measurement and when a change in the force profile was observed. The images were taken using The images were taken using a x40 objective lens, Ti-E microscope (Nikon Instruments, Melville, SCIeNTIFIC REPORTS | (2018) 8:9350 | DOI:10.1038/s41598-018-27548-1 NY, USA) and a optiMOS TM sCMOS Camera (Qimagimg, Surrey, Canada) (Sup Fig. 1). Each measurement was performed using at least two different bacterial cells and/or biofilm samples. In detail: B. subtilis WT on a glass: 3 bacteria, 117 force curves, adhesion frequency: 73.5% B. subtilis WT on a B. subtilis biofilm: 2 bacteria on 2 different biofilms, 118 force curves adhesion frequency: 85.6%, B. subtilis WT on S. mutans biofilm: 3 bacteria on 2 different biofilms, 53 force curves adhesion frequency: 100%, B. subtilis ∆tasA on glass: 3 bacteria, 46 force curves adhesion frequency: 61%, B. subtilis ∆tasA on S. mutans biofilm: 2 bacteria on one biofilm 100 force curves adhesion frequency: 51%.
Importantly, control measurements between a PDA coated cantilever and, glass, B. subtilis biofilm and S. mutans biofilm were taken (Supplementary Fig. S2).
AFM measurements Analysis. Analysis was conducted using JPK data processing software. The adhesion force was considered as the highest attraction force in the curve (largest negative force), and the detachment work energy was evaluated by calculating the area between the curve and the baseline. Curves not exhibiting adhesion events were discarded from the adhesion force and energy calculation.
Characterization of biofilm structure by CLSM. CLSM was used for visualization of deeper layers of the biofilm and quantification of B. subtilis cells in the dual-species biofilm. Dual-species biofilm was prepared as described above using B. subtilis cells tagged with GFP. Incorporation of dextran-associated EP was visualized by adding 1 µM Alexa Fluor 647-labeled dextran conjugate (molecular weight, 10,000 MW, Molecular Probes Inc.) to the growth media, during biofilm formation as described previously 38 .
The generated biofilms were washed carefully using PBS and visualized by Zeiss LSM510 CLS microscope (Carl Zeiss, Oberkochen, Germany). Three-dimensional images were constructed using Zen software (Carl Zeiss). At least three random fields were observed in five independent experiments. The amount of EP produced by S. mutans and viable B. subtilis cells was calculated as blue and green fluorescence intensity, respectively, using ImageJ software (http://rsb.info.nih.gov/ij). dual-species biofilm was tested using increased concentrations of dextranase 47,48 . Dextranase solution was prepared by dissolving 3 mg of lyophilized dextranase enzyme (Sigma Aldrich, St. Louise, Missouri, USA) in 1 ml of PBS. The dextranase solution was added to the dual-species formed biofilms in different amounts equal to: 156, 300, 600 and 900 units of enzyme. The generated biofilms were taken for visualization by CLSM as mentioned above. The amount of B. subtilis cells were calculated by fluorescent intensity using ImageJ software.

B. subtilis
Quantification of tasA and eps operons expression. To analyse the effect of S. mutans presence on tasA and eps expression, we used B. subtilis P tapA-lacZ or P eps-lacZ strains. B. subtilis cells were grown separately or with S. mutans, in BHI supplemented with 2% sucrose. To determine the effect ∆pH on tasA expression, 30% lactic acid solution (Merck & Co., Inc. Kenilworth, New Jersey, US) was used for the adjustment of pH of the medium to pH 6. Samples were collected every 12 h; normalization conducted using CFU on selective plates (streptomycin-LB agar plates). O.D was calculated according to 1 O.D (600 nm) = 1.5*10^8 CFU 49 . β-Galactosidase activity assay was conducted as described previously 50 .
Statistical analysis. The obtained data were statistically analysed using t-test. When needed ANOVA following post-hoc t-test with Bonferroni correction was conducted. All tests applied were two-tailed, and a p-value of 5% or less was considered statistically significant. Statistical analysis was performed using Microsoft Excel software.
Data availability. The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.