Efficient Biofilm-Based Fermentation Strategies by eDNA Formation for l-Proline Production with Corynebacterium glutamicum.

Biofilms could provide favorable conditions for the growth of cells during industrial fermentation. However, biofilm-immobilized fermentation has not yet been reported in Corynebacterium glutamicum (C. glutamicum), one of the main strains for amino acid production. This is mainly because C. glutamicum has a poor capability of adsorption onto materials or forming an extracellular polymeric substance (EPS). Here, an engineered strain, C. glutamicum Pro-ΔexeM, was created by removing the extracellular nuclease gene exeM, which effectively increased extracellular DNA (eDNA) in the EPS and cell adhesiveness onto carrier materials. In repeated-batch fermentation using the biofilm, l-proline production increased from 10.2 to 17.1 g/L. In summary, this research demonstrated that a synthetic C. glutamicum biofilm could be favorable for l-proline production, which could be extended to other industrial applications of C. glutamicum, and the strategy may also be applicable to the engineering of other strains.

Introduction

Biofilms, as microbial communities, are dynamic environments wherein cells propagate attached to organic or inorganic surfaces, as in dental plaque, food processing equipment, various water and ventilation pipes, medical equipment, etc.[1−3] Biofilms have some “social” attributes such as aggregate growth, adhesive medium, and quorum sensing.[4,5] Many strains on Earth can exist in the form of biofilms like Clostridium acetobutylicum, Escherichia coli (E. coli), Bacillus subtilis, and Saccharomyces cerevisiae.[6,7] The extracellular polymeric substance (EPS) matrix of biofilms mainly contains extracellular proteins, extracellular polysaccharides, and extracellular DNA (eDNA). These complex components determine the structure of biofilms and allow the bacteria in biofilms to adapt to a changing environment.[8,9] Recently, researchers have discovered that cells in the biofilm could grow steadily with high activity, providing favorable conditions for industrial fermentation.[10,11] Moreover, cells covered by biofilms could tolerate more hash conditions such as high osmotic pressure, oxygen limitations, and high cell density during fermentation.[12] Subsequently, biofilm-immobilized fermentation was proposed and applied in the fermentation industry by improving strain adsorption capacity.

During biofilm-immobilized fermentation, the biofilm cells can withstand high-speed agitation.[13] Furthermore, the biofilm cells that attach to the carrier surface could be renewed when fermentation broth is replaced with a fresh medium.[14] Due to high cell activity and the repetitive usage of cells in the biofilm-immobilization fermentation, seed culture can be eliminated, the lag phase can be reduced, and the production cycle could be shortened substantially.[15]

In biofilms, eDNA plays an important role in the connection among the carriers, cells, extracellular polysaccharides, and proteins.[16,17] eDNA was shown to have a great influence on the biofilm structure of Streptococcus mutans and Staphylococcus aureus.[18,19] Sumei Liao et al.[20] found that eDNA could be woven with extracellular polysaccharides to form an inseparable flocculent structure in Streptococcus mutans biofilms. In addition, eDNA could maintain the extracellular biofilm matrix[21,22] and acted as a “kite line” that restricted the outward release of the bacteria at the top of the biofilm structure. eDNA participates in the entire life cycle of biofilms from the initial stage of formation to the dissociation.[23] So, eDNA was not only the key to biofilm formation but also the main component maintaining biofilm structure.[24] Therefore, an increase in eDNA accumulation would be beneficial to biofilm formation. Extracellular nuclease is one of the factors responsible for the degradation of eDNA.[20] Researchers have shown that the removal of nuclease genes exeM and exeS promoted the accumulation of eDNA and biofilm formation in Shewanella oneidensis (S. oneidensis) MR-1.[25] On the other hand, many bacteria have a type IV secretion system (T4SS) that could secret DNA or DNA-protein complexes.[30,31] Within the T4SS, the VirB11 family proteins were the most important for the DNA and protein to go through.[26−28] Although there was no research on exeM or VirB11 in Corynebacterium glutamicum (C. glutamicum), in this study, we identified two genes highly homologous to exeM and VirB11 in C. glutamicum ATCC 13032-ProB, which produces l-proline.[29] Subsequently, we engineered the C. glutamicum ATCC 13032-ProB strain to enhance biofilm formation by removing exeM and overexpressing VirB11. Overall, this study represents a successful case for the development of biofilm-based fermentation under aerobic (nonstatic) industrial conditions for efficient biochemical production.

Results

Effect of Extracellular DNA on Biofilms

DNA fragments could be degraded by DNase I effectively without influencing the cell growth.[17] Thus, 10 μL of DNase I (5 U) was added to each well in order to reduce the DNA content within C. glutamicum ATCC 13032-ProB biofilms at various intervals. As expected, samples with inactivated DNase I showed a better biofilm formation ability, which indicated that the eDNA played an important role throughout the whole process of biofilm formation (Figure A). In addition, some researches verified the effects of DNase I on the biofilm matrix, which had been formed completely.[17] The cells in biofilms were stained with DAPI and detected by a CLSM (Figure B). The morphological structure of the biofilm with inactivated DNase I was tighter and more uniform (Figure B, left) compared with that of the biofilm that had added active DNase I (Figure B, right).

Figure 1

Effect of addition of DNase I on biofilm formation in the original strain. (A) Effect of DNase I on the biofilm amount during the fermentation process. DNase I (10 μL) (5 U) and 20 μL of seed culture were added to each well of a 96-well plate with 180 μL of the fermentation medium. Another group of samples with inactivated DNase I was used as a negative control. The biofilm amount was detected at the 12, 24, 36, 48, and 60 h time points. (B) CLSM images of biofilms, which were cultivated for 72 h at 30 °C with DNase I.

Extracellular secretions were observed, and the distance of two adjacent cells in the biofilm was less than 200 nm (Figure A). So, a length of 2000 bp DNA fragment (680 nm) would be sufficient to connect adjacent cells. Results showed that the biofilm formation with different concentrations (0.6–2.4 ng/μL) and lengths (2000–5000 bp) of added DNA was all stronger than that without the DNA addition (OD570 = 0.12). In particular, 0.6 ng/μL 5000 bp DNA gave the most apparent biofilm formation (Figure B).

Figure 2

EPS matrix of C. glutamicum ATCC 13032-ProB cells and quantitative analysis of the effect of DNA addition on biofilm formation in the original strain. (A) Distance of C. glutamicum ATCC 13032-ProB cells in the biofilm, as indicated by the red signs. The biofilm was formed on a cell slide in a 6-well plate after 72 h of cultivation at 30 °C. (B) Effects of added DNA fragments on biofilm formation. DNA fragments with different lengths (2000, 3000, 4000, or 5000 bp) and different final concentrations (0.6, 1.2, 1.8, and 2.4 μg/mL) were added at the beginning of cultivation. Experiments were carried out in 96-well plates, and biofilm was quantified after 72 h of incubation at 30 °C.

Extracellular DNA Content of Engineered Strains

PCR and sequencing results confirmed that the three recombinant strains in which the VirB11 gene was overexpressed (Pro-VirB11) or the exeM gene was knocked out (Pro-ΔexeM1, Pro-ΔexeM) were constructed successfully. The eDNA concentration of the original and three recombinant strains increased rapidly in the first 12 h and then increased steadily throughout the fermentation stage. At the end of the batch fermentation, the eDNA concentration of the original bacteria, Pro-ΔexeM1, Pro-ΔexeM, and Pro-VirB11 were 16.3, 23.2, 27.4, and 25.2 μg/mL, respectively. Pro-ΔexeM had the highest amount of extracellular DNA that was 68.1% higher than the original strain (Figure A).

Figure 3

Comparison of eDNA accumulation and biofilm formation between the recombinant strains and the original strain. The incubation time and conditions were the same as in Figure A. C. glutamicum ATCC 13032-ProB was abbreviated to 13032-ProB in the legends. (A) Differences in eDNA formation between the four strains during the cultivation process. (B) Biofilm formation ability of the four strains under different media. (C) Visible biofilm observed for the Pro-ΔexeM strain but not for the original strain (13032-ProB).

The 96-well plate experiment showed different biofilm formation abilities of the strains, as was detected by the crystal violet semi-quantitative method using the fermentation medium. The optical density from the crystal violet staining (which was an indicator of biofilm quantity) for Pro-ΔexeM was increased by 75% compared with the original strain (2.17 vs 1.24), while Pro-ΔexeM1 was increased by 41.9% (1.76 vs 1.24). Similar results were found in Pro-VirB11, which exhibited an increase of 41.1% (1.75 vs 1.24). The activation medium (LBG) and the seed medium were tried as well, and similar results were observed (Figure B).

The biofilm formed by the Pro-ΔexeM strain, which had the highest eDNA content, could be visibly observed (Figure C) after being cultured for 72 h in a fermentation medium in a 6-well plate, whereas the biofilm formation by the original strain could not be observed under such methods.

Biofilm Observation

Weak fluorescence was observed in the original strain by a DAPI fluorescent dye (Figure A), while relatively strong fluorescence was found in the recombinant strains Pro-ΔexeM1 (Figure B), Pro-ΔexeM (Figure C), and Pro-VirB11 (Figure D). This indicated that the biofilm in the original strain was scarce and scattered, whereas the biofilm of the recombinant strains was thick and abundant (Figure ).

Figure 4

Confocal laser scanning microscopy images of the biofilm formed by different strains. The biofilm was stained with a DAPI fluorescent dye. The incubation time and conditions were the same as Figure A. (A) C. glutamicum ATCC 13032-ProB, (B) Pro-ΔexeM1, (C) Pro-ΔexeM, and (D) Pro-VirB11.

DDAO staining and subsequent CLSM indicated that extracellular DNA surrounded the recombinant cells in the biofilm. The red staining of eDNA by DDAO in Pro-ΔexeM (Figure A-2) and Pro-VirB11 (Figure A-3) was significantly stronger and wider than the original strain (Figure A-1). An overlay of bright fields and dark fields showed that more cells were colocalized with clumps of eDNA (Figure A). A 3D recontruction of the eDNA staining of Pro-ΔexeM revealed that a large amount eDNAs formed an extensive three-dimensional complex structure filled with holes and extrusions (Figure B).

Figure 5

Confocal laser scanning microscopy images of three different strains stained with a DDAO fluorescent dye. The incubation time and conditions were the same as Figure A. (A) CLSM images of the eDNA in different strains at different fields. (A-1) C. glutamicum ATCC 13032-ProB, (A-2) Pro-ΔexeM, and (A-3) Pro-VirB11. (B) Three-dimensional image of eDNA in Pro-ΔexeM stained with a DDAO fluorescent dye.

Biofilm-Based Fermentation for Enhanced L-Proline Production

The exeM involved in eDNA accumulation did not affect the metabolic capacity of strains. Simultaneously, the Pro-ΔexeM strain showed a greater ability in its biofilm formation. Thus, the Pro-ΔexeM strain was chosen for immobilized repeated-batch fermentation.

l-Proline production in the first four batches had improved gradually in Pro-ΔexeM (Figure A). After the fourth batch, l-proline production was maintained at an average of 16.2 g/L, and l-proline productivity was kept at around 0.34 g/L/h, which was much higher compared with that of the original strain (Figure B) (0.34 g/L/h vs 0.14 g/L/h). Through a fermentation process of 420 h, l-proline production was increased by 66% in Pro-ΔexeM compared with that of the original strain (17.1 g/L vs 10.3 g/L). Furthermore, the fermentation period was shortened from 72 to 48 h. Therefore, the immobilized repeated-batch fermentation taking advantage of the biofilm formation in the exeM-knockout strain could enhance the l-proline concentration and productivity.

Figure 6

Biofilm-based immobilized fermentation. Polyurethane foam at 30 g/L was added as a biofilm carrier. At the end of each batch, the fermented broth was removed, and the biofilm carrier was left for repeated use in the next batch that was initiated by adding a fresh culture medium. (A) l-proline production by Pro-ΔexeM. (B) l-proline production by C. glutamicum ATCC 13032-ProB.

Discussion

C. glutamicum ATCC 13032-ProB was used to establish a biofilm-based fermentation for production of l-proline. This original strain formed less biofilms, which limited its industrial application.

Extracellular DNA for Enhanced Biofilm Formation

The original strains formed less biofilms because the extracellular DNase I hydrolyzed the eDNA during the biofilm formation process or because the biofilm could not form due to insufficient eDNA at the initial stage. In the culture medium, different lengths and concentrations of added DNA were all demonstrated to promote the formation of biofilms. This suggested that addition of exogenous DNA would affect biofilms positively. In particular, the results showed that a low dose (0.6 ng/μL) of longer DNA fragments (5000 bp) could be more beneficial for biofilm formation.

Molecular Modification of Genes

We identified the gene homologs of exeM in S. oneidensis MR-1 by protein blast from NCBI and searched the VirB11 gene in C. glutamicum ATCC 13032 from NCBI. The knockout of exeM and overexpression of VirB11 resulted in significant improvement in the formation of biofilms. It was uncertain that the knockout Pro-ΔexeM1 strain has removed one exeM gene or two exeM genes. However, we removed all exeM genes in the Pro-ΔexeM strain.

Furthermore, the Pro-ΔexeM strain produced the most eDNA and biofilms, suggesting that reducing the extracellular nucleases might be more effective than trying to promote the secretion of eDNA through VirB11.

Biofilm-Based Immobilized Fermentation

The biofilm-based immobilized fermentation strategy was widely applied in industrial production.[32] The Pro-ΔexeM strain with a better biofilm formation ability was selected for the immobilized fermentation of l-proline, and it achieved greater production levels compared to the other strains. The Pro-ΔexeM would be of great value for the immobilized fermentation of l-proline, and the addition of a biofilm carrier was also indispensable. The carrier could fix bacteria on its surface and support biofilm development, but it had to be suitable for oxygen and mass transfer during the cell growth process.[33] The bacteria cells in the liquid fermentation broth were dramatically decreased, which was beneficial for product separation. Compared with free-cell fermentation, the carrier with attached biofilms served as an immobilized biocatalyst that could be reused again and again, with a cell renewal ability. This eliminated the need for seed culture. However, biofilm reactors will be complicated by the biofilm carrier. Modification and screening of carriers should also be carried out in the future. This study would also provide a reference for developing more biochemical-producing processes based on C. glutamicum biofilms and applying eDNA to other biofilm-forming bacteria.

Materials and Methods

Genes, Strains, and Plasmids

The research strain C. glutamicum ATCC 13032-ProB is a derivative from C. glutamicum ATCC 13032. The genes, strains, and plasmids used in this work are listed in Table , and the primers are in Table .

Table 1

Bacterial Strains, Plasmids and Genes Used in This Study

strains	relevant characteristics	source
C. glutamicum ATCC 13032-ProBa	original strain, l-proline producer	Prof. Sheng Yang
E. coli DH5α-pXMJ19	Cm resistance	stored in our lab
E. coli DH5α-pk18mobsacB	Kan resistance	stored in our lab
Pro-ΔexeM1	C. glutamicum ATCC 13032-ProB with deleted part exeM	this study
Pro-ΔexeM	C. glutamicum ATCC 13032-ProB with deleted exeM	this study
Pro-VirB11	C. glutamicum ATCC 13032-ProB harboring plasmid pXMJ19-VirB11	this study
Plasmids
pk18mobsacB	Kan resistance	this study
pXMJ19	Cm resistance	this study
pk18mobsacB-exeM	pk18mobsacB with ΔexeM from C. glutamicum ATCC 13032-ProB	this study
pXMJ19-VirB11	pXMJ19 with VirB11 from C. glutamicum ATCC 13032-ProB	this study
Genes
exeM	CGL-RS12940, ExeM/NucH family extracellular endonuclease	NCBI
VirB11	Cg10301, predicted ATPases involved in pili and flagella biosynthesis, VirB11 family	NCBI

A gift from Prof. Sheng Yang (Institute of Plant Physiology & Ecology, CAS, Shanghai, China).

Table 2

Main Primers Used in This Study

primers	relevant characteristics
exeM1-up-F	TCACTTCGCGACCTCAACTC
exeM1-up-R	agtggcagggcggggcgtaaACAGACAATTTGTTGCTGGT
exeM1-down-F	ccagtgatttttttctccatgcatTGGCAGCCTCGACCAC
exeM1-down-R	CAGACCTGCGAGCCGACGGC
Cm-exeM1-F	ACCAGCAACAAATTGTCTGTttacgccccgccctgccact
Cm-exeM1-R	GTGGTCGAGGCTGCCAATGCatggagaaaaaaatcactggatataccaccgttgat
exeM-up-F	aattcgagctcggtacccggggatccTCACTTCGCGACCTCAACTCCC
exeM-up-R	ACAGACAATTTGTTGCTGGTCTCAGG
exeM-down-F	CCTGAGACCAGCAACAAATTGTCTGTGCATTGGCAGCCTCGACCACG
exeM-down-R	gcctgcaggtcgactctagaggatccCAGACCTGCGAGCCGACGG
VirB11-F	gcctgcaggtcgactctagaggatccATGACTGACATTGATCTGGTGGTGGAA
VirB11-R	aattcgagctcggtacccggggatccCTAGGGCATAAACCATGCCTCTTCG
Y-exeM-F	TCACTTCGCGACCTCAACTC
Y-exeM-R	CAGACCTGCGAGCCGACGGC
Y-VirB11-F	aattaagcttgcatgcctgcaggt
Y-VirB11-R	atcggcgctacggcgtttca

Gene Source

The exeM gene in S. oneidensis MR-1 was used to blast proteins in C. glutamicum ATCC 13032 by NCBI. An ExeM/NucH family extracellular endonuclease (CGL-RS12940) was found and was named exeM in this study. On the other hand, a gene (Cg10301) predicted to encode a VirB11 family ATPase, involved in pili and flagella biosynthesis, was named VirB11 in this study.

Strain Engineering

One or two copies of the exeM gene in the parental strain were removed by the long flanking homology region-PCR (LFH-PCR) method[34] to obtain a strain named Pro-ΔexeM1. The PCR primers were designed by using the National Center for Biotechnology Information (NCBI) gene sequence database and SnapGene design. Briefly, a PCR-generated Chloramphenicol resistance marker was used as a knock-in DNA fragment. The Chloramphenicol resistance marker consisted of a Chloramphenicol resistance sequence in homologous regions (around 1500 bp) flanking the target locus.

Complete removal of exeM was achieved by using a pk18mobsacB plasmid. The resulting strain was named Pro-ΔexeM.[35] Briefly, a PCR-generated Kanamycin resistance marker was used as a knock-in DNA fragment. The Kanamycin resistance marker consisted of a Kanamycin resistance sequence in plasmid pk18mobsacB and homologous regions (around 1500 bp) flanking the target locus. The knock-in component, which was named pk18mobsacB-exeM, was transformed into strain C. glutamicum ATCC 13032-ProB using a Bio-Rad electroporation system set at 1.8 kV and 25 mF with a 200 Ohm pulse controller.

The VirB11 gene was amplified from the genome DNA of C. glutamicum ATCC 13032-ProB and ligated to the overexpression plasmid pXMJ19 (BamH I) to obtain the plasmid pXMJ19-VirB11. The VirB11 gene and plasmid pXMJ19 (with restriction enzyme BamH I) were ligated by using the ClonExpress II One Step Cloning Kit C112-01 (Vazyme, Nanjing, China), resulting in a plasmid pXMJ19-VirB11. The final engineered strain was named Pro-VirB11 with Chloramphenicol resistance for screening.

Media and Growth Conditions

C. glutamicum ATCC 13032-ProB, Pro-ΔexeM1, Pro-ΔexeM, and Pro-VirB11 were cultured in an LBG medium containing 5 g/L yeast extract, 10 g/L tryptone, 10 g/L NaCl, and 10 g/L glucose. The agar plates were prepared by adding 1.5% (w/v) agar into LBG media. The seed medium contained 25 g/L glucose, 17.5 g/L corn steep liquor, 5 g/L (NH4)2SO4, 0.5 g/L MgSO4·7H2O, 1 g/L KH2PO4, and 2 g/L urea. The fermentation medium contained 100 g/L glucose, 20 g/L corn steep liquor, 30 g/L (NH4)2SO4, 0.4 g/L MgSO4·7H2O, 1.2 g/L KH2PO4, 2 g/L urea, and 15 g/L CaCO3, and 2.4 g/L urea was added to the medium every 12 h. Acetic acid was added to the medium to adjust its initial pH level to 7.2. Kanamycin (50 mg/mL), Chloramphenicol (50 mg/mL), sucrose (10 g/L), isopropyl-β-d-thiogalactopyranoside (IPTG) (0.5 mM) or l-arabinose (100 mM) was added as required. The fermentation culture was grown at 30 °C with an agitation speed of 220 rpm.

For the free-cell fermentation, the flasks were inoculated with 6 to 9% (v/v) seed culture and then shaken for 72 h at 220 rpm in 30 °C. The culture was sampled every 12 h to monitor the glucose and l-proline.

The same conditions were used for immobilized repeated-batch fermentation with 30 g/L polyurethane foam as a biofilm carrier. At the end of the first batch, the fermented broth was removed from the flask, and the carrier that was covered by biofilms was left for the second batch. After adding a fresh culture medium, the second batch was initiated under the same conditions. The subsequent batches were operated in the same way as above.

Carrier Preparation

A novel porous polyurethane foam was prepared in the laboratory. The carrier had a density of 0.63 g/cm3 with a pore diameter of 0.2 to 0.4 mm sheared to a size of 5 mm × 5 mm × 5 mm. This carrier was pretreated using the previously reported method.[36] The carrier was rinsed in 1 M NaOH and then 1 M HCl before being washed with sterile water until the pH value reached 7.0. The carrier was sterilized at 121 °C for 15 min before use.

Analytical Methods

l-Proline was measured by HPLC (Agilent 1260 series; Hewlett-Packard, Palo Alto, CA, USA) with a UV detector, using a Sepax AAA ion exclusion column (250 × 4.6 mm; Bio-Rad Laboratories, Hercules, CA, USA), with sodium acetate and 80% acetonitrile as the mobile phase (0.6 mL/min) at 39 °C. In this study, we used triethylamine acetonitrile, phenyl isothiocyanate acetonitrile, and n-hexane to make l-proline become weak-polar derivatives that could be detected by HPLC. Glucose was measured by a refractive index detector by using an Aminex HPX-87H column (300 × 7.8 mm), with 5 mM H2SO4 as the mobile phase (1 mL/min) at 55°C.[37]

After the cell culture matured, the medium was discarded, and the 96-well plate was rinsed twice with PBS (1%) and prepared for washing the free cells. Methanol (200 μL) was used for fixing the biofilm for 10 min. Then, methanol was poured out, and the 96-well plate was dried at room temperature. The biofilm was stained with 200 μL of 2% crystal violet for 10 min at room temperature. After that, all wells were rinsed gently with PBS three times and patted lightly on absorbent paper towels, and then, the plate was left to dry at room temperature or inside an incubator at 30 °C. For semi-quantitative purposes, 200 μL of 30% glacial acetic acid was added, and the wells were incubated by agitating gently for 30 min. The absorbance of crystal violet was an indicator of biofilm quantity, and this was measured at 570 nm by using a multiscan spectrum (SpectraMax iD5; Molecular Devices, USA).[38]

eDNA was extracted by the Gram-positive bacteria genome extraction kit (TaKaRa MiniBEST Bacteria Genomic DNA Extraction Kit Ver.3.0). To quantify the eDNA amount, denatured fish sperm DNA solution (Sangon Biotech B548210, 5 mg/mL) was combined with a Hoechst 33258 fluorescent dye and used to draw a standard curve.[39,40] The maximum excitation wavelength and maximum emission wavelength were determined in the microplate reader. The standard curve equation was y = 2411.3x + 3059.8 (R2 = 0.9915), and the linear range was 0.1–1.0 μg/mL. Samples that contained 1 mL of extracted eDNA were diluted 50 times with an equal volume of TNE buffer (pH 7.4) and a Hoechst 33258 fluorescent dye. These were mixed evenly and shielded from light at room temperature for 5 min. Each sample (200 μL) was immediately added to a black 96-well plate, and then, the DNA concentration was detected by the microplate reader.

A confocal laser scanning microscope (CLSM) (Leica TCS SP5II, Wetzlar, Germany) was used to visualize the distribution of biofilms and the content of eDNA. The cell slide was gently removed from the fermentation broth in the 6-well plate. PBS (1%) was used to wash the slides twice; then, 2.5% glutaraldehyde was used to fix the biofilm for 30 min; then, PBS (1%) was used to wash the biofilm three times. 4′,6-Diamidino-2-phenylindole (DAPI) (Sigma, St. Louis, MO, USA)[41] and 7-hydroxy-9H-(1,3-dichloro-9,9-dimethylacridin-2-one) (DDAO) (AAT Bioquest, CA, USA)[42] fluorescent dyes (0.2 μg/mL) were added. They were left to settle and stain the nuclei and eDNA for 20 min; then, PBS (1%) was used to wash them three times. An anti-fluorescence quencher (50 μL) was applied on the cell slide after the cell slide was dried at room temperature in the dark. Finally, the cell slide glass was sealed with nail polish and observed under the CLSM immediately.

Two fluorescent dyes were used to visualize the biofilm morphology. A DAPI fluorescent dye was used to stain the nucleus of the strains that were observed. A DDAO fluorescent dye was used to stain the eDNA of the three strains. DAPI was able to penetrate the cell membranes and bound to double-stranded DNA with strong bluish fluorescence. DDAO was unable to penetrate the cell membranes but could bind to extracellular double-stranded DNA with strong red fluorescence.

A field emission scanning electron microscope (FESEM) (SEM JSM-6360LV, Jeol Ltd., Japan) was used to visualize the biofilm morphology. The cell slide with attached biofilms was taken out from the 6-well plate and washed three times with PBS (1%). Glutaraldehyde (2.5%) was used to fix the biofilm at 4 °C for 12 h, and then, PBS (1%) was used to wash the cell slide. The cell slide was placed in the refrigerator at −80 °C overnight; then, it was dehydrated by a vacuum freeze-drying device (Labconco Corporation, Fort Scott, Kansas, USA) and coated with gold–palladium before it was placed in an FESEM for analysis.