Transcriptomic Analysis Reveals the Role of tmRNA on Biofilm Formation in Bacillus subtilis

Bacillus strains are widely distributed in terrestrial and marine environments, and some of them are used as biocontrol organisms for their biofilm-formation ability. In Bacillus subtilis, biofilm formation is fine-tuned by a complex network, a clear understanding of which still requires study. In bacteria, tmRNA, encoded by the ssrA gene, catalyzes trans-translation that can rescue ribosomes stalled on mRNA transcripts lacking a functional stop codon. tmRNA also affects physiological bioprocesses in some bacteria. In this study, we constructed a ssrA mutant in B. subtilis and found that the biofilm formation in the ssrA mutant was largely impaired. Moreover, we isolated a biofilm-formation suppressor of ssrA, in which the biofilm formation was restored to a level even stronger than that in the wild type. We further performed RNAseq assays with the wild type, ssrA mutant, and suppressor of ssrA for comparisons of their transcriptomes. By analyzing the transcriptomic data, we predicted the possible functions of some differentially expressed genes (DEGs) in the tmRNA regulation of biofilm formation in B. subtilis. Finally, we found that the overexpression of two DEGs, acoA and yhjR, could restore the biofilm formation in the ssrA mutant, indicating that AcoA and YhjR were immediate regulators involved in the tmRNA regulatory web controlling biofilm formation in B. subtilis. Our data can improve the knowledge about the molecular network involved in Bacillus biofilm formation and provide new targets for manipulation of Bacillus biofilms for future investigation.


Introduction
Biofilms are communities of surface-associated microbial cells enclosed by an extracellular matrix. Biofilms are the predominant mode of microbial growth in natural environments. Microbial biofilms are intimately related to human life. On the one hand, most human infections caused by microbial biofilms are notoriously challenging to treat because the microorganisms are embedded in a protective matrix of extracellular polymeric substances, for instance, exopolysaccharides (EPS) [1]. Apparently, the protection of biofilms confers multidrug resistance and immune evasion capabilities on bacteria, which makes them highly pathogenic and difficult to eradicate [2]. Biofilms formed by bacterial pathogens lead to prolonged infections and increased mortality rate among patients, which has always been a hot topic in medicine, in the past several decades [3]. On the other hand, biofilms produced by some bacteria can be beneficial to humans. They have shown great potential in wastewater treatment and production of microbial fuels [4,5]. The ability to

Construction of Strains
To delete the ssrA gene in B. subtilis strains, we used the CRISPR-Cas9 system harbored in the plasmid vector pJOE8999, which is a shuttle plasmid containing a temperaturesensitive replicon for B. subtilis [34]. First, we subcloned the gRNA sequence for ssrA in pJOE8999 at the site of Bsa I, and obtained the resulting plasmid pYX101. Then, the DNA sequences located upstream and downstream of ssrA were PCR-amplified from the genomic DNA of strain NCIB 3610 using the primer pairs ssrA-U-F/R and ssrA-D-F/R, respectively. These two fragments were subsequently sutured together via SOE-PCR to construct the fragment ssrA-UD. The ssrA-UD fragment was subcloned in pYX101 at site Sfi I, and the resulting plasmid pYX102 was then transformed into JM110, to be demethylated as a new plasmid pYX103. Plasmid pYX103 was subsequently introduced into the target strain requiring a ssrA deletion via transformation, and the transformants were selected as Kan R colonies. The verified Kan R colonies were further grown at 45 • C to remove pYX103 that originated from the temperature-sensitive plasmid pJOE8999. The pYX103-free strains were screened as Kan S colonies, and whether the disruption of ssrA via double-crossover homologous recombination occurred in these colonies during the incubation at 45 • C was tested using PCR with the primer pair ssrA-verify-F/R. Mutants with deleted genes acoA, yhjR, sinR, and skfA in B. subtilis were constructed in the same manner as that described for the ssrA mutant. To complement the ssrA mutation, plasmid pNW33N [35], a replicative and high-copy plasmid for Bacillus, was used as the vector carrying the wild-type ssrA gene. The sequence of ssrA, including its promoter region, was PCR-amplified from the genomic DNA of NCIB 3610 using the primer pairs ssrA-33N-F/R and obtained as a fragment flanked by the sequences of Pst I and BamH I. The amplified ssrA fragment was then subcloned in pNW33N between sites Pst I and BamH I, and the constructed plasmid was designated as pYX201 and transformed into JM110 for demethylation. The demethylated recombinant plasmid, pYX202, was finally transformed into the target host strains as described earlier [34]. In addition, to overexpress genes, we subcloned the fusion of a strong promoter pG and gene of interest in the pNW33N plasmid using the ClonExpress ® MultiS One Step Cloning Kit (Vazyme Biotech Co., Ltd., Nanjing, China), according to the manufacturer's instructions. All the primers used in this study are listed in Table S2.

Biofilm Formation Assays
Strains were grown in LB to an optical density at 600 nm (OD 600 ) of 1.0. For the colony assay, the LB cultures were 10 times diluted with sterile deionized water and spotted on the MSgg agar (1.5% agar), and the seeded agar plates were incubated at 30 • C for different times. For the pellicle assay, 1.5 µL of the diluted culture was added to 1.5 mL LBMG medium contained within a well of a 24-well microtiter plate, and the microtiter plates were incubated at 30 • C for the indicated times.

Swarming Motility Assay
To test swarming motility, single colonies of the strains grown on regular LB plates (1.5% agar) were picked and spotted onto the soft agar medium supplemented with LB broth and 0.65% agar. The seeded soft agar plates were incubated at 37 • C for 6 h, for observation of swarming zones.

Isolation of the Swarming Suppressors of ssrA in B. subtilis
The ssrA mutant was grown on regular LB plates and its single colonies were picked and spotted on the soft agar plates (the same as described above). The seeded plates were incubated at 37 • C till obvious swarming zones were observed. Bacteria from different areas of the outer rings of the swarming zones were picked and streaked on regular LB plates, for further purification of the suppressors of ssrA.

Confocal Laser Scanning Microscopy
The 48-h incubated submerged biofilms in microtiter plates were observed using a LSM710-Meta confocal laser scanning microscope (ZEISS, Jena, Germany). For the observation of biofilms in B. subtilis strains, cells were fluorescently stained in green with the nucleic acid marker SYTO ® 9 (Invitrogen by Thermo Fisher Scientific, Waltham, MA, USA), at the concentration of 5 µM (diluted in PBS). After a 20-min incubation in the dark at 37 • C to enable fluorescent labeling of the bacteria, the plate was mounted on the motorized stage of the confocal microscope. The biofilms were scanned at excitation wavelengths of 488 nm (argon laser; 3% intensity), with emission wavelengths collected from 493 to 550 nm for the SYTO ® 9 fluorescence. Three-dimensional (3D) projections of the biofilm structures were reconstructed using the Easy 3D function of the IMARIS software (Bitplane, Zürich, Switzerland) from the xyz image series.

Quantification of the Average Thickness and Biovolume of Biofilms
The average thickness and biovolumes of biofilms were calculated using COM-STAT [36]. The average biovolume was measured and calculated using results from three independent experiments, and the average thickness of biofilms was calculated using results from three independent experiments.

RNA Isolation
In order to select a suitable sampling time point, the growth conditions of WT, TM, and MB in the LBMG medium contained in 24-well microtiter plates were analyzed and the time point 36 h was selected. The cells at the selected time point were collected and their total RNA was extracted using the E.Z.N.A. ® Bacterial RNA Kit (OMEGA, Norcross, GA, USA), according to the manufacturer's instructions. The RIN numbers and gel electropherogram of the RNA samples are shown in the Supplementary Materials.

Transcriptome Analysis
The library construction and Illumina sequencing were performed at Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). A total amount of 3 µg RNA per sample was used as input material for the RNA sample preparations. mRNA was purified from total RNA using probes to remove rRNA. The purified mRNA was fragmented and used as templates for cDNA synthesis. The cDNA libraries were sequenced using Illumina HiSeq™  genome in the  sample of WT, 7,124,644 reads in the sample of TM, and 7,462,953 reads in the sample of MB. The differentially expressed genes (DEGs) were determined between WT and TM, TM and MB, and WT and MB with the standards of padj ≤0.05, and fold change |log 2 Ratio| > 1. The KEGG database (http://www.genome.jp/kegg/, accessed on 1 December 2020) was used to classify the genes with significant differential expression based on their functions. KOBAS software was used to test the statistical enrichment of differential expression genes in the KEGG pathways. In addition, we performed a Quantitative Real-time PCR experiment (Takara.RR430B TB Green Fast qPCR Mix) for the target genes to verify the reliability of the transcriptome data (see the Supplementary Materials).

Sporulation Assay
The sporulation assay was based on methods published in previous literature with some modifications according to the experimental requirements [37]. The samples were collected under the same conditions as described for the RNA isolation and serially diluted with sterile physiological saline. The collected cultures were well resuspended, centrifuged at 4500 rpm for 5 min, and then the supernatant was discarded. The harvested samples were washed with 500 µL physiological saline, and finally resuspended in physiological saline. The resuspended samples were subsequently subjected to mild sonication (20% amplitude, 1 s on, 1 s off for 5 s total) to liberate bacterial cells from the matrix. A 10 µL sub-sample of each cell sample was incubated for 20 min at 80 • C (heat-resistant) to kill vegetative cells. Then the samples were resuspended with 300 µL physiological saline, which was plated on LB agar plates and incubated for 12 h at 37 • C. The number of spores produced was determined as the number of colonies grown on the LB agar plates, for each sample. Meanwhile, 10 µL of each cell sample was directly resuspended with 300 µL physiological saline and were diluted (10,000 times) for plating. These samples were used as a control for the high-temperature treated samples.

Extraction of Extracellular Polysaccharide (EPS)
Colonies of the WT NCIB 3610 strain grown in the MSgg agar plates were crushed and transferred into a triangular flask. Sterile water was added to the flask for dissolution of the colonies, followed by incubation at 50 • C for 1.5-2 h, for extraction of the EPS. Then, the incubated mixture was centrifuged at 8000× g for 15 min, and the supernatant of the extract was collected. Next, 30% alcohol was added to the supernatant, which was subsequently precipitated at 4 • C for overnight. The supernatant was collected via centrifugation and further precipitated at 4 • C overnight, by adding 70% alcohol. The precipitate obtained was the extracted EPS. The precooled EPS samples were freeze-dried by a vacuum freeze-drying device for 24 h. The final white powdery substance was Bacillus crude EPS, which was weighed and quantified.

tmRNA Positively Affects Biofilm Formation in B. subtilis
To study the effect of tmRNA on B. subtilis physiology, the ssrA deletion mutant in the type strain NCIB 3610 was constructed by using the CRISPR-Cas9 system ( Figure S1). The biofilm formation in the wild-type (WT) and the ssrA mutant (TM) was investigated. The results showed that there was not a significant difference in biofilm formation between WT and TM when the strains were grown in MSgg medium ( Figure S2a). However, in LB medium, TM produced obviously less pellicles than WT. The WT strain could form intact biofilms in LBM, while the TM strain could not ( Figure S2a). When LBMG was used, the phenotype appeared to be stronger: WT started forming pellicles after 24-h incubation, whereas TM formed very little; WT formed more matured biofilms with wrinkles after 48-h incubation, whereas the TM only formed smooth or even broken biofilms at the same point ( Figures S2a and 1a). Thus, the LBMG medium was selected to grow bacteria for the following experiments required to be performed in liquid medium. The biofilm formation of WT and TM was also tested using different media plates. WT could produce biofilms with more wrinkled structures than TM during growth in all tested media plates ( Figure S2b). Given that the MSgg medium agar plates produced the largest differences in biofilm-formation phenotypes, this medium was used to grow strains for observation of biofilms on solid surfaces (Figures 1a and S2b).
To further confirm the effect of tmRNA on biofilm formation, the complementary strain (C∆ssrA) was constructed and the biofilm-production defect was largely restored in C∆ssrA (Figures 1a and S2c,d). Moreover, the biofilm architectures were analyzed using Confocal Laser Scanning Microscope (CLSM). Focusing on the sheets of cells attached to the surface, there were obviously a larger number of cells observed for WT than that for TM ( Figure 1b). The reconstructed images and 3D models of the confocal series showed that the biofilms produced by TM were much weaker and contained less protruded structures than those produced by WT (Figure 1b). The phenotypes in the surface-attached cells and biofilm formation observed for C∆ssrA resembled those of WT ( Figure 1b). Furthermore, the thickness and biovolume of biofilms formed by the strains described above was quantified. The average thicknesses of biofilms produced by WT, TM, and C∆ssrA were 101 µm, 60 µm, and 104 µm, respectively. The average biovolume calculated for WT, TM, and C∆ssrA were 4.834 × 10 3 µm 3 , 2.478 × 10 3 µm 3 , and 4.875 × 10 3 µm 3 , respectively ( Figure 1c). All these results suggest that tmRNA plays an important role in biofilm initiation and maturation in B. subtilis. between WT and TM when the strains were grown in MSgg medium ( Figure S2a). However, in LB medium, TM produced obviously less pellicles than WT. The WT strain could form intact biofilms in LBM, while the TM strain could not ( Figure S2a). When LBMG was used, the phenotype appeared to be stronger: WT started forming pellicles after 24-h incubation, whereas TM formed very little; WT formed more matured biofilms with wrinkles after 48-h incubation, whereas the TM only formed smooth or even broken biofilms at the same point ( Figures S2a and 1a). Thus, the LBMG medium was selected to grow bacteria for the following experiments required to be performed in liquid medium. The biofilm formation of WT and TM was also tested using different media plates. WT could produce biofilms with more wrinkled structures than TM during growth in all tested media plates ( Figure S2b). Given that the MSgg medium agar plates produced the largest differences in biofilm-formation phenotypes, this medium was used to grow strains for observation of biofilms on solid surfaces (Figures 1a and S2b). To further confirm the effect of tmRNA on biofilm formation, the complementary strain (CΔssrA) was constructed and the biofilm-production defect was largely restored in CΔssrA (Figures 1a and S2c,d). Moreover, the biofilm architectures were analyzed using Confocal Laser Scanning Microscope (CLSM). Focusing on the sheets of cells attached to the surface, there were obviously a larger number of cells observed for WT than that for TM ( Figure 1b). The reconstructed images and 3D models of the confocal series showed that the biofilms produced by TM were much weaker and contained less protruded structures than those produced by WT (Figure 1b). The phenotypes in the surface-attached cells and biofilm formation observed for CΔssrA resembled those of WT (Figure 1b). Furthermore, the thickness and biovolume of biofilms formed by the strains described above was quantified. The average thicknesses of biofilms produced by WT, TM, and CΔssrA were 101 μm, 60 μm, and 104 μm, respectively. The average biovolume calculated for WT, TM, and CΔssrA were 4.834 × 10 3 μm 3 , 2.478 × 10 3 μm 3 , and 4.875 × 10 3 μm 3 , respectively ( Figure  1c). All these results suggest that tmRNA plays an important role in biofilm initiation and maturation in B. subtilis.

Isolation of a Suppressor of ssrA That Cured the TM Biofilm-Formation Defect
In view of the association of biofilm formation with swarming motility in B. subtilis [38], in this study, the swarming motility in colonies of the WT strain showed obvious swarming zones, and yet colonies of the TM strain showed almost no swarming zones ( Figure 2a). This result indicates that lacking tmRNA caused dramatically diminished swarming motility in B. subtilis. It is likely that certain factors involved in the tmRNA regulation of biofilm formation also affect swarming motility in B. subtilis.

Isolation of a Suppressor of ssrA That Cured the TM Biofilm-Formation Defect
In view of the association of biofilm formation with swarming motility in B. subtilis [38], in this study, the swarming motility in colonies of the WT strain showed obvious swarming zones, and yet colonies of the TM strain showed almost no swarming zones ( Figure 2a). This result indicates that lacking tmRNA caused dramatically diminished swarming motility in B. subtilis. It is likely that certain factors involved in the tmRNA regulation of biofilm formation also affect swarming motility in B. subtilis.
It has been described that swarming suppressor mutants for low-motile B. subtilis strains can be isolated via prolonged incubation of the bacteria on soft agar plates [39]. Here, one of the strains obtained (designated as tm3MB (MB)) had a not only restored but reinforced biofilm-formation phenotype (Figures 2b and S3). The biofilms of MB were much thicker than those of WT, the average thickness of biofilms was 153 µm (MB) and 101 µm (WT) and the average biovolume calculated was 6.185 × 10 3 µm 3 (MB) and 4.834 × 10 3 µm 3 (WT) (Figure 2b  It has been described that swarming suppressor mutants for low-motile B. subtilis strains can be isolated via prolonged incubation of the bacteria on soft agar plates [39]. Here, one of the strains obtained (designated as tm3MB (MB)) had a not only restored but reinforced biofilm-formation phenotype (Figures 2b and S3). The biofilms of MB were much thicker than those of WT, the average thickness of biofilms was 153 μm (MB) and 101 μm (WT) and the average biovolume calculated was 6.185 × 10 3 μm 3 (MB) and 4.834 × 10 3 μm 3 (WT) (Figure 2b,c). All these results indicated that the suppressor mutant (MB)

Transcriptome Analysis among the WT, TM, and MB Strain
To investigate whether the tmRNA mutation and suppressor mutation(s) in MB caused any altered gene expression in the parent strain, an RNAseq analysis of WT, TM, and MB was performed. The quality of the RNA samples is shown in Figure S4. The results show that nearly 40% of the genes across the genome were differentially expressed in TM, compared with WT (Figure 3a), suggesting a major role of tmRNA in genetic regulation in B. subtilis. Furthermore, compared with TM, there are 1537 genes which had altered expression in MB (Figure 4a), indicating that the spontaneous suppressor mutation(s) in MB resulted in great reprogramming of gene expression in TM. Out of these 1537 genes, there are 1234 whose expression was also altered when comparing TM with WT, suggesting that most of the gene-expression differences found in MB were tmRNA related ( Figure S5). In addition, 178 genes were found differentially expressed in all the three comparisons ( Figure S5). Further study of the differentially expressed genes (DEGs) found in TM and MB might provide us valuable information regarding the effect of tmRNA on B. subtilis physiology, especially biofilm formation. biofilm formation in MB and are there novel unknown factors involved in biofilm formation in B. subtilis? To answer these questions, more in-depth experiments are required.

Transcriptome Analysis among the WT, TM, and MB Strain
To investigate whether the tmRNA mutation and suppressor mutation(s) in MB caused any altered gene expression in the parent strain, an RNAseq analysis of WT, TM, and MB was performed. The quality of the RNA samples is shown in Figure S4. The results show that nearly 40% of the genes across the genome were differentially expressed in TM, compared with WT (Figure 3a), suggesting a major role of tmRNA in genetic regulation in B. subtilis. Furthermore, compared with TM, there are 1537 genes which had altered expression in MB (Figure 4a), indicating that the spontaneous suppressor mutation(s) in MB resulted in great reprogramming of gene expression in TM. Out of these 1537 genes, there are 1234 whose expression was also altered when comparing TM with WT, suggesting that most of the gene-expression differences found in MB were tmRNA related ( Figure  S5). In addition, 178 genes were found differentially expressed in all the three comparisons ( Figure S5). Further study of the differentially expressed genes (DEGs) found in TM and MB might provide us valuable information regarding the effect of tmRNA on B. subtilis physiology, especially biofilm formation.  . Analysis of the gene-expression differences between TM and WT. (a) A volcano plot showing the DEGs found in TM compared with WT. DEGs were defined as the genes whose expression was up-regulated (pink dots) or down-regulated (blue dots) with a fold change of two or more (p ≤ 0.05), when TM was compared with WT. Black dots represent genes without a significant change in expression. (b) Profiling of the DEGs via using the annotation of KEGG. (c) Grouping the DEGs by the indicated metabolic pathways. (d) Changes in the expression of the known biofilm−formation genes found as DEGs.

Gene-Expression Differences between TM and WT
There were 756 genes up-regulated and 992 genes down-regulated with a fold change of two or more comparing TM with WT ( Figure 3a). These genes were classified into 20 major cellular processes based on the annotation of KEGG (Figure 3b), and further grouped into three major metabolic pathways related to biofilm formation, including

Gene-Expression Differences between TM and WT
There were 756 genes up-regulated and 992 genes down-regulated with a fold change of two or more comparing TM with WT (Figure 3a). These genes were classified into 20 major cellular processes based on the annotation of KEGG (Figure 3b), and further grouped into three major metabolic pathways related to biofilm formation, including chemotaxis, flagellar assembly and quorum sensing (Figure 3c). Overall, the results showed that tmRNA mediated the expression of genes involved in many cellular processes with biosynthesis of secondary metabolites, microbial metabolism in diverse environments, and ABC transporters being the most mediated ones (Figure 3b). Surprisingly, the expression of known genes required for chemotaxis and flagellar assembly was either not altered or even increased in TM (Figure 3c), indicating that tmRNA affected B. subtilis swarming motility and biofilm formation without repressing the transcription of flagellar genes or through flagellar-independent mechanism(s). Moreover, tmRNA could either positively or negatively regulate the expression of some genes involved in quorum sensing (Figure 3b,c). According to the DEGs, tmRNA regulated the transcription of genes encoding for known factors controlling biofilm formation in a highly unusual way (Figure 3d). For instance, tmRNA could not only repress the transcription of some known activators, such as SigX, Abh, EpsA, and EpsE, but also some repressors, such as AbrB, of B. subtilis biofilm formation (Figure 3d). In addition, tmRNA was required for the expression of comER, the product of which is an activator for both biofilm formation and sporulation in B. subtilis (Figure 3d). All these data suggest that tmRNA functions as an important regulator controlling many cellular processes in B. subtills, possibly via a highly complex network.
Among these DEGs, the genes whose expression was greatly affected (top 20) by tmRNA were analyzed (Tables S3 and S4). The genes with an expression pattern that was either consistent with or opposite to the pattern of biofilm formation in WT, TM, and MB were focused on. For instance, the expression of gene ytcB and acoA was decreased in TM and yet elevated in MB, compared with the WT (Table S4), which was regulated in a similar manner as that found for biofilm formation. These data suggested that AcoA and YtcB are potential intermediate activators within the tmRNA regulatory cascade(s) of biofilm formation in B. subtilis. It has been described that during biofilm development, the acetoin utilization genes (acoABCL operon) expression was among the most significant upregulation [40]. In addition, acoA and acoB can be coregulated by interconnected webs of transcriptional factors, including Fnr, AcoR, SigL, and CcpA [41]. On the other hand, the product of ytcB is a putative UDP-glucose epimerase, which is expressed late during sporulation in the mother cell, and its expression can be inhibited by GerE and σ K [42]. These results indicated the tmRNA has effects on acetoin metabolism and UDP-glucose modification, the study of which might provide important information elucidating the mechanism(s) for the tmRNA regulation of biofilm formation in B. subtilis.

Gene-Expression Differences between MB and TM
Comparing MB with TM, there were 974 genes up-regulated whereas 563 genes were down-regulated with a fold change of two or more (Figure 4a). In addition, these genes were classified into 20 major cellular processes based on the annotation of KEGG (Figure 4b), and further grouped into three major metabolic pathways related to biofilm formation ( Figure 4c). These data showed that compared with TM, the most affected cellular processes in MB were the biosynthesis of secondary metabolites, microbial metabolism in diverse environments, and ABC transporters, which was the same as that found in the comparison between TM and WT (Figure 4b). In addition, the distribution of up-regulated and down-regulated genes in the shown bioprocesses comparing MB with TM was almost opposite to that in the TM with WT comparison (Figures 3b and 4b), suggesting that the gene-expression difference caused by tmRNA could be largely restored via the suppressor mutation(s) in the MB strain. Still, the DEGs involved in metabolisms of chemotaxis, flagellar biosynthesis, quorum sensing, and biofilm formation did not provide us sufficient information to unveil the mechanisms underneath the physiological phenotypes found in MB and TM (Figure 4c,d).
We also analyzed the most up-regulated genes (Table S5) and down-regulated genes (Table S6) in the MB with TM comparison. The data showed that all the genes up-regulated in TM, compared with WT, were down-regulated in MB, compared with TM, and vice versa (Tables S5 and S6). This result further confirmed the tight connection between tmRNA and the suppressor mutation(s) in MB. Moreover, the genes of acoA, ytcB, acoL, gerPA, yngJ, and ydgA were down-regulated in TM and yet overexpressed in MB, compared with WT, the expression pattern of which resembled the biofilm-formation pattern observed in these three strains (Table S5). In contrast, hxlB and rocA were expressed totally in a reverse manner, in WT, TM, and MB (Table S6). Among these genes, acoA and ytcB have been described above, and acoL belongs to the acoABCL operon. gerPA encodes a germination protein GerPA, and this gene is expressed during sporulation in the mother cell. Expression of the σ K -dependent gerP operon is repressed by GerE [43,44]. The yngJ gene is located in the yngJIHGFE gene cluster in B. subtilis, which is expressed during sporulation in the mother cell. According to the profiling based on Swiss-Prot, YngJ is a homologue of acyl-CoA dehydrogenase [45]. The ydgA gene encodes a conserved hypothetical protein, and the expression of this gene is regulated by σ K [46]. In addition to AcoA and YtcB, AcoL, GerPA, YngJ, and YdgA could be potential activators of biofilm formation, involved in the regulatory cascade(s) of tmRNA. On the other hand, gene products of hxlB and rocA are also noteworthy because they could be potential repressors of biofilm formation within the tmRNA regulatory web. HxlB is a 6-phospho-3-hexuloisomerase involved in the ribulose monophosphate pathway in B. subtilis [47]. RocA is a L-glutamate gamma-semialdehyde dehydrogenase involved in proline metabolism [48]. Clearly, all these results provided us more information about how tmRNA might affect B. subtilis physiology.

Gene-Expression Differences between MB and WT
Comparing MB with WT, there were 179 genes up-regulated and 307 genes downregulated with a fold change of two or more ( Figure 5a). These genes were classified into 20 major cellular processes based on the annotation of KEGG (Figure 5b). Clearly, the number of DEGs found in this comparison was much smaller than that found in the other two comparisons described above (Figures 3a and 4a). This might be because most of the gene-expression differences resulted from the tmRNA deletion were restored to a WT level in the MB strain. In fact, more than half of the DEGs shown in Figure 5 were also found in the comparison between TM and WT. Thus, the profiling results for the DEGs in the MB vs. WT comparison were similar to those found in the TM vs. WT comparison (Figures 3b and 5b). Notably, there are bioprocesses, such as ribosome metabolism, which seemed to be uniquely caused by the suppressor mutation(s) in MB rather than the tmRNA deletion (Figure 5b). Grouping the DEGs in the MB vs. WT comparison into biofilmrelated metabolic pathways did not provide sufficiently precise information (Figure 5c). Yet, profiling of these DEGs based on known biofilm-formation genes showed that, in addition to tapA, the other two genes sipW and tasA included in the same operon were up-regulated in MB, which was not observed in the TM vs. WT comparison. Further study of this operon might provide information elucidating the mechanism of the over-produced biofilms in MB ( Figure 2).
Again, we analyzed the most up-regulated genes (Table S7) and down-regulated genes (Table S8) in MB, compared with WT. We still focused on the genes with an expression pattern that was similar to acoA or hxlB, as described in the comparison between TM and MB. Two more genes, yhjR and yxbB, possessed an acoA-similar expression pattern in strains of WT, TM, and MB. The gene yhjR encodes a spore coat protein with the function of protecting the spores, and its localization depends on SafA [49]. The yxbB gene encodes a putative S-adenosyl-L-methionine-dependent methyltransferase, whose function has not been studied, according to the information based on the SubtiWiki database [50]. Gene yxbB is located in the asnH operon comprising yxbB, yxbA, yxnB, asnH, and yxaM, which is induced dramatically during the transition from exponential to stationary phase when the cells are grown in rich sporulation medium [51]. These data suggest that YhjR and YxbB might also be intermediates of the tmRNA regulatory pathway(s) for biofilm formation in B. subtilis. Again, we analyzed the most up-regulated genes (Table S7) and down-regulated genes (Table S8) in MB, compared with WT. We still focused on the genes with an expression pattern that was similar to acoA or hxlB, as described in the comparison between TM and MB. Two more genes, yhjR and yxbB, possessed an acoA-similar expression pattern in strains of WT, TM, and MB. The gene yhjR encodes a spore coat protein with the function of protecting the spores, and its localization depends on SafA [49]. The yxbB gene encodes a putative S-adenosyl-L-methionine-dependent methyltransferase, whose function has not been studied, according to the information based on the SubtiWiki database [50]. Gene yxbB is located in the asnH operon comprising yxbB, yxbA, yxnB, asnH, and yxaM, which is induced dramatically during the transition from exponential to stationary phase when the cells are grown in rich sporulation medium [51]. These data suggest that YhjR and YxbB might also be intermediates of the tmRNA regulatory pathway(s) for biofilm formation in B. subtilis.

Elucidating the Roles of Known Biofilm-Formation Factors in the tmRNA Regulation of Biofilm in B. subtilis
Given that the extracellular matrix is an important structure for biofilm complexes, we extracted the EPS from WT, and added the obtained EPS into TM cultures. As shown in Figure 6, the biofilm-formation ability of TM was restored almost to the WT level, when 100 µg mL −1 EPS was added to the cultures. This result indicates that tmRNA affects biofilm formation through positively regulating EPS production in B. subtilis. The transcriptome results regarding EPS production were then checked precisely. Unfortunately, as described earlier for Figure 3d, the known EPS-related genes were regulated in an unclear manner. For instance, both the structure genes, such as epsA and epsE, and the repressor gene abrB were up-regulated in TM (Figure 3d and Table S9), which appeared to be paradoxical and not consistent with the biofilm-formation phenotypes we found here. In MB, the epsA expression was restored to a WT level, whereas the epsE expression was not altered significantly, and the epsB expression was increased obviously, compared with TM (Table S9), which contributed nothing to explain the tmRNA effect on EPS production. Moreover, some other extracellular-matrix activator genes such as tapA, sigX, and abh were also up-regulated in TM (Figure 3d and Table S9), making the tmRNA regulatory network more complex. Interestingly, the whole operon of tapA-sipW-tasA was overexpressed in MB, compared with WT, which was not observed when comparing TM with WT ( Figure 5d and Table S9). Given that the tapA-sipW-tasA operon encodes the amyloid fiber of the extracellular matrix, it is worth investigating whether the over-produced biofilms were (partially) related to the overexpression of this operon in MB. However, this phenotype specifically found in MB is likely to be tmRNA-independent. All these data suggested that tmRNA might affect the known EPS-related genes at post-transcriptional levels or it controlled unknown pathway(s) producing the extracellular matrix in B. subtilis. manner. For instance, both the structure genes, such as epsA and epsE, and the repressor gene abrB were up-regulated in TM (Figure 3d and Table S9), which appeared to be paradoxical and not consistent with the biofilm-formation phenotypes we found here. In MB, the epsA expression was restored to a WT level, whereas the epsE expression was not altered significantly, and the epsB expression was increased obviously, compared with TM (Table S9), which contributed nothing to explain the tmRNA effect on EPS production. Moreover, some other extracellular-matrix activator genes such as tapA, sigX, and abh were also up-regulated in TM (Figure 3d and Table S9), making the tmRNA regulatory network more complex. Interestingly, the whole operon of tapA-sipW-tasA was overexpressed in MB, compared with WT, which was not observed when comparing TM with WT ( Figure 5d and Table S9). Given that the tapA-sipW-tasA operon encodes the amyloid fiber of the extracellular matrix, it is worth investigating whether the over-produced biofilms were (partially) related to the overexpression of this operon in MB. However, this phenotype specifically found in MB is likely to be tmRNA-independent. All these data suggested that tmRNA might affect the known EPS-related genes at post-transcriptional levels or it controlled unknown pathway(s) producing the extracellular matrix in B. subtilis. Figure 6. The effect of EPS on TM biofilm formation. The strains of WT and TM were grown in the LBMG liquid medium, supplemented with or without 100 μg mL −1 EPS, and observed for biofilm formation. The scale represents 5 mm.
It is known that SinR represses the expression of matrix genes in B. subtilis [52]. Here, the transcriptome results show that, compared with WT, expression of sinR was increased by approximately 50% in TM and decreased by approximately 45% in MB. which was consistent with the biofilm-formation pattern in these three strains. Subsequently, we deleted the sinR gene in TM to test whether the biofilm-formation phenotype could be restored. Unfortunately, our results show that the deletion of sinR did not rescue the biofilm defect in TM ( Figure S6), suggesting the existence of other pathways involved in the tmRNA regulation of biofilm formation in B. subtilis.

Initial Characterization of the Unknown Pathways Involved in the tmRNA Regulation of Biofilm Formation in B. subtilis
To better understand how tmRNA functions in the complex regulatory network for biofilm formation in B. subtilis, the effects of several genes (acoA, acoL, gerPA, ydgA, yhjR, Figure 6. The effect of EPS on TM biofilm formation. The strains of WT and TM were grown in the LBMG liquid medium, supplemented with or without 100 µg mL −1 EPS, and observed for biofilm formation. The scale represents 5 mm. It is known that SinR represses the expression of matrix genes in B. subtilis [52]. Here, the transcriptome results show that, compared with WT, expression of sinR was increased by approximately 50% in TM and decreased by approximately 45% in MB. which was consistent with the biofilm-formation pattern in these three strains. Subsequently, we deleted the sinR gene in TM to test whether the biofilm-formation phenotype could be restored. Unfortunately, our results show that the deletion of sinR did not rescue the biofilm defect in TM ( Figure S6), suggesting the existence of other pathways involved in the tmRNA regulation of biofilm formation in B. subtilis.

Initial Characterization of the Unknown Pathways Involved in the tmRNA Regulation of Biofilm Formation in B. subtilis
To better understand how tmRNA functions in the complex regulatory network for biofilm formation in B. subtilis, the effects of several genes (acoA, acoL, gerPA, ydgA, yhjR, yngJ, ytcB, and yxbB) whose transcription was highly regulated by tmRNA were investigated. Here, all these genes are over-expressed in TM, but only overexpression of acoA and yhjR could cause restored phenotypes in biofilm formation in the TM strain (Figures 7a and S7). yngJ, ytcB, and yxbB) whose transcription was highly regulated by tmRNA were investigated. Here, all these genes are over-expressed in TM, but only overexpression of acoA and yhjR could cause restored phenotypes in biofilm formation in the TM strain (Figures  7a and S7). As shown in Figure 7a, the strain of TM/acoA + possessed stronger biofilm formation in liquid cultures, and this phenotype was not obvious on agar plates. However, biofilms produced by the TM/acoA + strain were still weaker than WT. The reconstruction of confocal images showed that the overexpressed acoA enabled TM to produce intact biofilms rather than broken biofilms (Figure 7a). In addition, the thickness and biomass of the TM/acoA + pellicles were similar to that of the WT pellicles (Figure 7b). These data indicate that the overexpression of acoA could partially restore the biofilm formation in TM. To further verify the role of AcoA in biofilm formation, the acoA gene was deleted in WT and MB. It is found that when grown in the liquid medium, the ∆acoA mutants in both WT and MB showed weaker biofilm production than the parent strains, after incubation for 24 h. However, this phenotype was not observed for the ∆acoA mutants after incubation for 48 h ( Figure S8a). These results suggest that the ∆acoA mutation could cause delayed biofilm formation in B. subtilis grown in the LBMG liquid medium. No obvious phenotypes in biofilm formation were found when the strains were grown on the MSgg agar plates ( Figure S8b).
The overexpression of yhjR appeared to restore the biofilm production in TM more efficiently. As shown in Figure 7a, on the MSgg agar plate, the recovery of the biofilm formation in TM/yhjR + was quite obvious, and in the LBMG liquid medium, the pellicles formed by strain TM/yhjR + were comparable to those of the WT strain. The reconstruction of confocal images showed that the yhjR overexpression enabled TM to fully restore the biofilm formation (Figure 7a), and the thickness and biomass of the TM/yhjR + pellicles were similar to that of the WT pellicles ( Figure 7b). These data indicate that the overexpression of yhjR could largely complement the biofilm formation in TM. However, we failed to describe the biofilm-formation phenotypes in the yhjR mutants because different isolates, which were supposed to be genetically identical, of the yhjR mutant in WT showed various phenotypes in biofilm formation. More experiments are required to explain and resolve this problem in our future studies. In summary, all these data suggest that the tmRNA regulation of biofilm might involve pathways of acetoin metabolism (AcoA) and sporulation (YhjR) in B. subtilis. More experiments in the future are needed to explain the AcoA and YhjR functions in the same or independent regulatory cascade originated from tmRNA.

Sporulation in WT, TM, and TM Derivates
Given sporulation is highly related to biofilm formation in B. subtilis [53], sporulation assays with the strains of WT, TM, MB, TM/acoA + and TM/yhjR + were performed. As shown in Figure 8, TM produced dramatically fewer spores than WT, after high-temperature treatments at 80 • C. In contrast, MB produced more spores than WT after treatments at 80 • C ( Figure 8). These phenotypes in sporulation were reminiscent of those in biofilm formation found in TM and MB, compared with WT. Moreover, we found that the overexpression of acoA could not restore the sporulation defect in TM, whereas the overexpression of yhjR could increase the sporulation in TM even to the MB level (Figure 8), which confirmed the physiological function of YhjR in sporulation. Our data indicated that YhjR might be a key factor connecting tmRNA and sporulation and might be used as a checkpoint for functional sporulation affecting the downstream bioprocesses such as biofilm formation in B. subtilis. The transcriptome data also supported the tight relationship between tmRNA and sporulation, given that most of the known sporulation-related genes were differentially expressed in TM compared with WT, and most of these DEGs showed restored expression in MB (Table S10). Obviously, studies in the molecular mechanism by which tmRNA mediates sporulation in B. subtilis will shed light on how tmRNA affects the physiology in this organism.

Discussion
In this study, we investigated the effect of tmRNA on biofilm formation in B. subtilis, which has not been described in Bacillus before. First, we constructed the ssrA-clean-deletion mutant (TM), using the CRISPR-Cas9 system in B. subtilis NCIB 3610. The results showed that TM produced remarkably less biofilms than WT (Figure 1), and showed a severe defect in swarming compared with WT ( Figure 2a). We then isolated swarming suppressors of ssrA and found that one of these suppressors, tm3MB (MB), possessed not only a restored but even a reinforced biofilm-formation phenotype compared with WT ( Figures S3 and 2b,c). Thus, we considered the MB strain as a biofilm-formation suppressor of ssrA and performed various assays with these strains, to elucidate the possible mechanism(s) underlying the tmRNA regulation of biofilm formation in B. subtilis.
We analyzed the gene-expression differences among WT, TM, and MB by comparing the transcriptome-sequencing results of these strains. Compared with WT, a large number of genes were differentially expressed in TM, and the DEGs covers many cellular processes including biofilm-related ones ( Figure 3). Compared with TM, the spontaneous suppressor mutation(s) in MB caused differential expression of many genes in this strain ( Figure 4). Most of the DEGs found in the TM vs. WT comparison had restored expression to the WT level in the MB strain, indicating that the suppressor mutation(s) in MB were highly related to tmRNA in B. subtilis. Compared with WT, MB possessed fewer DEGs than TM (Figure 5a), and most of the DEGs were the ones resulting from the tmRNA deletion ( Figure 5b). However, there were DEGs that appeared to be MB-specific, such as the genes grouped into the ribosome metabolism and amyloid fiber biosynthesis ( Figure  5b,d). Unfortunately, simply grouping the DEGs based on KEGG annotation and biofilmrelated bioprocesses did not provide sufficiently precise information to target pathway(s) mediating the tmRNA regulation of biofilm formation. Thus, we focused on particular DEGs that were most regulated in the comparisons. Finally, two sets of DEGs with special expression patterns were found in WT, TM, and MB. The first set includes acoA, acoL, gerPA, ydgA, yhjR, yngJ, ytcB, and yxbB, whose expression is regulated in a similar manner by which biofilm formation is regulated based on the comparison of WT, TM, and MB (Tables S4, S5 and S7). The second set includes hxlB and rocA, whose expression is regulated in the opposite manner by which biofilm formation is regulated based on the comparison of WT, TM, and MB (Tables S3 and S6). These genes could be considered as potential intermediate activators and repressors of biofilm formation in the tmRNA regulatory cascade(s) in B. subtilis.

Discussion
In this study, we investigated the effect of tmRNA on biofilm formation in B. subtilis, which has not been described in Bacillus before. First, we constructed the ssrA-clean-deletion mutant (TM), using the CRISPR-Cas9 system in B. subtilis NCIB 3610. The results showed that TM produced remarkably less biofilms than WT (Figure 1), and showed a severe defect in swarming compared with WT ( Figure 2a). We then isolated swarming suppressors of ssrA and found that one of these suppressors, tm3MB (MB), possessed not only a restored but even a reinforced biofilm-formation phenotype compared with WT ( Figures S3 and 2b,c). Thus, we considered the MB strain as a biofilm-formation suppressor of ssrA and performed various assays with these strains, to elucidate the possible mechanism(s) underlying the tmRNA regulation of biofilm formation in B. subtilis.
We analyzed the gene-expression differences among WT, TM, and MB by comparing the transcriptome-sequencing results of these strains. Compared with WT, a large number of genes were differentially expressed in TM, and the DEGs covers many cellular processes including biofilm-related ones ( Figure 3). Compared with TM, the spontaneous suppressor mutation(s) in MB caused differential expression of many genes in this strain ( Figure 4). Most of the DEGs found in the TM vs. WT comparison had restored expression to the WT level in the MB strain, indicating that the suppressor mutation(s) in MB were highly related to tmRNA in B. subtilis. Compared with WT, MB possessed fewer DEGs than TM (Figure 5a), and most of the DEGs were the ones resulting from the tmRNA deletion ( Figure 5b). However, there were DEGs that appeared to be MB-specific, such as the genes grouped into the ribosome metabolism and amyloid fiber biosynthesis (Figure 5b,d). Unfortunately, simply grouping the DEGs based on KEGG annotation and biofilm-related bioprocesses did not provide sufficiently precise information to target pathway(s) mediating the tmRNA regulation of biofilm formation. Thus, we focused on particular DEGs that were most regulated in the comparisons. Finally, two sets of DEGs with special expression patterns were found in WT, TM, and MB. The first set includes acoA, acoL, gerPA, ydgA, yhjR, yngJ, ytcB, and yxbB, whose expression is regulated in a similar manner by which biofilm formation is regulated based on the comparison of WT, TM, and MB (Tables S4, S5 and S7). The second set includes hxlB and rocA, whose expression is regulated in the opposite manner by which biofilm formation is regulated based on the comparison of WT, TM, and MB (Tables S3 and S6). These genes could be considered as potential intermediate activators and repressors of biofilm formation in the tmRNA regulatory cascade(s) in B. subtilis.
Furthermore, the roles of known biofilm-related factors EPS and SinR were tested. The results showed that addition of EPS could restore the biofilm formation in TM to a WT level (Figure 6), suggesting that tmRNA regulated biofilm formation by affecting EPS production. However, the genes involved in EPS or other extracellular matrix structures were regulated in an unclear manner that could not explain the phenotypes found for the biofilm formation and EPS production in TM, compared with WT (Table S9). This might be because the EPS-related genes were differentially expressed at post-transcriptional levels or that there were unknown tmRNA-controlled processes dedicated to EPS production in B. subtilis. In addition, our data showed that even the expression of sinR was slightly up-regulated in TM and down-regulated in MB, compared with WT. Knocking out the sinR gene in TM did not restore its biofilm-formation defect ( Figure S6), which indicates that the tmRNA regulation of biofilm formation in B. subtilis is SinR-independent. Thus, studies seeking for novel pathways involved in the tmRNA regulatory network mediating biofilm formation are essential.
To elucidate the unknown factors connecting tmRNA and biofilm formation in B. subtilis, we tested the effects of several genes (acoA, acoL, gerPA, ydgA, yhjR, yngJ, ytcB, and yxbB) whose special expression pattern has been described above. The data show that overexpressing acoA could partially restore the biofilm defect in TM (Figure 7), and the acoA mutants showed delayed biofilm formation compared with the parent strains ( Figure S8), confirming the role of AcoA in the tmRNA regulation of biofilm formation. Moreover, the overexpression of yhjR largely restored the biofilm-formation defect in TM (Figure 7), indicating its role in connecting tmRNA and biofilm formation in B. subtilis. The acoABCL operon encodes the acetoin dehydrogenase complex, and the yhjR gene encodes spore coat protein. Thus, tmRNA might control biofilm formation through different pathways including acetoin metabolism and sporulation. In fact, we observed that sporulation was obviously affected in TM and MB, compared with WT, in a similar manner to that observed for biofilm formation (Figure 8). The overexpression of yhjR caused greatly restored sporulation in TM (Figure 8), which further confirmed the role of YhjR in the tmRNA regulatory network. It is possible that YhjR was used as a checkpoint for successful sporulation which positively affected biofilm formation in B. subtilis. However, unveiling the full web comprising tmRNA, acetoin metabolism, sporulation, and biofilm formation definitely requires more investigation in our future studies.
Despite the initial findings for the molecular mechanism by which tmRNA regulates biofilm formation in B. subtilis, we are far from a clear understanding of the whole tm-RNA regulatory network in this bacterium. To accomplish this, we can use the following strategies in the future studies: (1) Given that tmRNA catalyzes trans-translation that is dedicated to rescuing stalled ribosomes, it might affect the proteome more directly than the transcriptome in bacteria. Thus, we can investigate the proteome patterns in WT, TM, MB, and other TM derivates, which might provide us more precise information about the tmRNA regulation. (2) We have identified the suppressor mutations in MB, which were listed in Figure S9 and Table S11, and further analyses of these mutations are required because the existence of them restored the biofilm-formation defect resulting from the tmRNA deletion. In addition, isolation and characterization of more biofilm suppressors of ssrA could be useful. (3) We should continue investigating the transcriptome results obtained in this study, especially the DEGs with an expression pattern similar or opposite to that for acoA and yhjR in WT, TM, and MB ( Figure S10), because these findings about the roles of AcoA and YhjR in the tmRNA effect on biofilm formation provided us insights into tmRNA and its regulation in B. subtilis. (4) It has been reported that some compounds, such as surfactin, bacillomycin D, and bacillaene, can promote biofilm formation in B. subtilis [54][55][56], and we found that there were some compound-related genes, such as the ones encoding sporulation killing factor, bacillaene, bacilysin, plipastatin, and CDPS, down-regulated in TM, compared with WT, and the expression of most of these genes was restored in MB to a WT level (Table S12 and Figure S11). These results indicate that the reduction of biofilm formation in TM might be related to compound production, but that the enhancement of biofilm formation in MB might not be related to compound production. Furthermore, we found that the deletion of skfA that encodes a sporulation killing factor caused impaired biofilm formation in WT and MB ( Figure S12), which partially supported our hypothesis about the relationship between compound production and biofilm formation, which were both regulated by tmRNA. Thus, investigation of the functions of particular compounds in the tmRNA regulation of biofilm formation will also be a promising research direction.

Conclusions
In conclusion, we demonstrated that the tmRNA deletion caused a biofilm-formation defect in B. subtills NCIB 3610. We successfully isolated a biofilm suppressor of ssrA, designated as MB, that could form thicker and more complex biofilms even than the WT strain. By analyzing the transcriptome data for WT, TM, and MB, we finally located two genes, acoA and yhjR, that could positively affect biofilm formation in B. subtilis, given that the overexpression of these genes caused restored phenotypes in biofilm formation in TM. We also conducted an initial characterization of the close relationship among tmRNA, sporulation, and biofilm formation in B. subtilis. Our data contribute more knowledge to the molecular regulatory network of biofilm formation in B. subtilis and may provide novel targets for better utilizing and modifying biofilm formation in Bacillus strains.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/microorganisms10071338/s1, Figure S1: Schematic illustration of the deletion for ssrA (left) and PCR validation (right); Figure S2: Comparison of the biofilm formation among WT, TM, and C∆ssrA; Figure S3: Isolation of suppressor(s) of ssrA and comparison of the biofilm formation among WT, TM, and isolated strains; Figure S4: The quality of the RNA samples and qPCR validation of transcriptome data. Figure S5: A Venn diagram showing the distribution of DEGs in comparisons WT vs. TM, TM vs. MB, and WT vs. MB; Figure S6: Comparison of the biofilm formation among WT, TM, and TM/∆sinR; Figure S7: Comparison of the biofilm formation among WT/33N, TM/33N, and overexpression strains; Figure S8: Comparison of the biofilm formation among WT, WT/∆acoA, MB, and MB/∆acoA; Figure S9: Diagrams showing where the suppressor mutations were located and the consequences of these mutations in MB; Figure S10: Trend analysis of DEGs; Figure S11: Transcriptome analysis of genes related to compound biosynthesis pathways; Figure S12: Comparison of the biofilm formation among WT, WT/∆skfA, MB, and MB/∆sfkA.; Table S1: Strains and plasmids used in this study; Table S2: Primers used in this work; Table S3: The most up-regulated genes comparing TM with WT; Table S4: The most down-regulated genes comparing TM with WT; Table S5: The most up-regulated genes comparing MB with TM; Table S6: The most down-regulated genes comparing MB with TM; Table S7: The most up-regulated genes comparing MB with WT; Table S8: The most down-regulated genes comparing MB with WT; Table S9: The biofilm-related genes differentially expressed among strains WT, TM, and MB; Table S10: The sporulation-related genes differentially expressed among strains WT, TM, and MB; Table S11: The expression of genes spontaneously mutated in the MB strain; Table S12: The expression of genes related to secondary metabiotic products in strains WT, TM, and MB. Reference [57] is cited in the Supplementary Materials.
Author Contributions: S.X. designed and performed experiments, analyzed data and wrote the paper. Q.C. performed experiments and analyzed data. Z.L., J.C. and P.Y. analyzed data and helped write the paper. Y.X. and B.L. designed experiments, analyzed data and wrote the paper. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The RNA sequence data have been deposited in the GEO database of NCBI with the accession code GSE199151. The whole-genome sequencing data that support the findings of this study have been deposited in the Sequence Read Archive with the project number PRJNA817865.