Author response: A toxin-mediated policing system in Bacillus optimizes division of labor via penalizing cheater-like nonproducers

Abstract

Full text Figures and data Side by side Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Division of labor, where subpopulations perform complementary tasks simultaneously within an assembly, characterizes major evolutionary transitions of cooperation in certain cases. Currently, the mechanism and significance of mediating the interaction between different cell types during the division of labor, remain largely unknown. Here, we investigated the molecular mechanism and ecological function of a policing system for optimizing the division of labor in Bacillus velezensis SQR9. During biofilm formation, cells differentiated into the extracellular matrix (ECM)-producers and cheater-like nonproducers. ECM-producers were also active in the biosynthesis of genomic island-governed toxic bacillunoic acids (BAs) and self-resistance; while the nonproducers were sensitive to this antibiotic and could be partially eliminated. Spo0A was identified to be the co-regulator for triggering both ECM production and BAs synthesis/immunity. Besides its well-known regulation of ECM secretion, Spo0A activates acetyl-CoA carboxylase to produce malonyl-CoA, which is essential for BAs biosynthesis, thereby stimulating BAs production and self-immunity. Finally, the policing system not only excluded ECM-nonproducing cheater-like individuals but also improved the production of other public goods such as protease and siderophore, consequently, enhancing the population stability and ecological fitness under stress conditions and in the rhizosphere. This study provides insights into our understanding of the maintenance and evolution of microbial cooperation. Editor's evaluation This manuscript reports notable findings regarding the potential for self-policing and a division of labor among biofilm-inhabiting Bacillus cells. Overall, this work is robust in its use of various techniques and provides solid insights into the intersections of well-understood regulatory controls and the suppression of cheaters. Colleagues interested in microbial social interactions should find this study's narrative about the internal mediation of cell differentiation valuable. https://doi.org/10.7554/eLife.84743.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Cooperative interactions are not restricted to complex, higher organisms, but are also prevalent among microbial communities in many contexts (Kehe et al., 2021; Rakoff-Nahoum et al., 2016; Wenseleers and Ratnieks, 2006). Both natural selection and game theory predict that cooperative systems are vulnerable to non-cooperative cheaters that exploit the benefit, such as public goods including extracellular enzymes (Chen et al., 2019), siderophore (Griffin et al., 2004), or biofilm matrix (Dragoš and Kovács, 2017; Vlamakis et al., 2013) since these selfish individuals enjoy the common resources without paying their cost (Hardin, 1968; Martin et al., 2020; West et al., 2006). Intriguingly, cooperation principally survives cheating during evolutionary history (Travisano and Velicer, 2004), and a couple of mechanisms have been proposed to play significant roles in maintaining cooperation by preventing cheater invasion (Özkaya et al., 2017; Travisano and Velicer, 2004). These strategies mainly include kin selection/discrimination (Özkaya et al., 2017; Diggle et al., 2007; McNally et al., 2017), facultative cooperation regulated by a quorum-sensing (QS) system (Allen et al., 2016), or nutrient fitness cost (Sexton and Schuster, 2017), coupling production of public and private goods (Dandekar et al., 2012), punishment of cheating individuals by cooperator-produced antibiotics (García-Contreras et al., 2015; Wang et al., 2015), partial privatization of public goods under certain conditions (Jin et al., 2018; Otto et al., 2020), and spatial structuring (van Gestel et al., 2014). In general, the emergence of multiple sanction mechanisms is a consequence of natural selection, which suppresses social cheaters and enhances the altruistic behavior, thereby maintaining microbial community stability and improving their adaptation in different niches (Özkaya et al., 2017). In certain cases, microbial cooperation involves the division of labor, where subpopulations of cells are specialized to perform different tasks (Dragoš et al., 2018a; Strassmann and Queller, 2011). Division of labor requires three basic conditions: individuals exhibit different tasks (phenotypic variation); some individuals carry out cooperative tasks that benefit other individuals (cooperation); all individuals gain an inclusive fitness benefit from the interaction (adaptation) (Dragoš et al., 2018b; West and Cooper, 2016). For instance, Bacillus subtilis colony will phenotypically differentiate into surfactin-producing and matrix-producing cells during sliding motility, where the surfactin reduces the friction between cells and their substrate, while the matrix assembles into van Gogh bundles that drive the migration (Jordi et al., 2015). Another typical case is in early-stage biofilms, an extracellular matrix (ECM)-the enclosed multicellular community that sustains bacterial survival in diverse natural environments; it is known that B. subtilis cells can differentiate into motile cells and matrix-producing cells during biofilm formation (Vlamakis et al., 2008; Chai et al., 2008; Shank and Kolter, 2011; López and Kolter, 2010; López et al., 2009c; López et al., 2009a; van Gestel et al., 2015; Vlamakis et al., 2013; Kearns, 2008). The advantage of the division of labor is to efficiently integrate distinct cellular activities, thereby endowing a community with higher fitness than undifferentiated clones (West and Cooper, 2016; Zhang et al., 2016). Importantly, efficient division of labor relies on elaborate coordination of cell differentiation (Cremer et al., 2019; López et al., 2009b; Lord et al., 2019). In relative to the subpopulation producing a certain public good (e.g. cells producing ECM during biofilm formation), the nonproducing cells that can also enjoy this common good, actually become the ‘cheater-like’ individuals to some extent (although they may provide other contributions to the community) (Claessen et al., 2014; Otto et al., 2020; West and Cooper, 2016). Therefore, regulating the proportion of each cell type and alignment of interests, is important for maintaining the stability and fitness of the division of labor (West and Cooper, 2016), while an unbalanced cell differentiation will reduce the population productivity and even cause a collapse of the division of labor (Dragoš et al., 2018a). Despite the knowledge of pathways controlling cell differentiation in microbes, little is known about how the different cell types interact with each other and the fitness consequences of their interaction (van Gestel et al., 2015). Although a few studies have investigated the overlap between public goods production and cell cannibalism (González-Pastor et al., 2003; López et al., 2009c), as well as matrix privatization (Otto et al., 2020) during cell differentiation, the molecular mechanism involved in coordinating the cheater-like individuals in the division of labor, as well as the ecological significance of the policing system in regulating population stability and fitness, remain unclear. Accordingly, lacking these knowledge limits our understanding of cooperation and altruism within microbial social communities. Bacillus velezensis SQR9 (formerly B. amyloliquefaciens SQR9) is a well-studied beneficial rhizobacterium that forms robust and highly structured biofilms on the air-liquid interface and plant roots (Qiu et al., 2014; Xu et al., 2019a; Xu et al., 2013; Cao et al., 2011). Strain SQR9 harbors a novel genomic island 3 (GI3) consisting of four operons, where the second, third, and fourth operons are responsible for the production of the novel branched-chain fatty acids, BAs, while the first operon encodes an ABC transporter to export toxic BAs for self-immunity (Wang et al., 2019). Production of toxic BAs was proved to occur in the subfraction of cells with the self-immunity ability induced by BAs during biofilm formation, where the nonproducing siblings will be lysed by BAs (Huang et al., 2021; Wang et al., 2019). Based on the manifestation that the BA-mediated cannibalism enhanced the biofilm formation of strain SQR9, we hypothesized the ECM and BAs synthesis can be co-regulated to restrain the cheater-like individuals that don’t produce ECM, thereby optimizing the division of labor and altruistic behavior. Using a combination of single-cell tracking techniques, molecular approaches, and ecological evaluation, we demonstrated that ECM and BAs production are coordinated in the same subpopulation by the same regulator during biofilm formation, which enforces punishment of the cheater-like nonproducers to maintain community stabilization; also this genomic island-governed policing system is significant to promote community fitness in various conditions. Results Coordinated production of ECM and autotoxin BAs punishes cheater-like nonproducers in the B. velezensis SQR9 community Bacillus cells in early-stage biofilms are known to contain specialized groups as motile cells and matrix-producing cells (Kearns, 2008; van Gestel et al., 2015; Vlamakis et al., 2008). We hypothesized that secretion of cannibal toxin BAs can eliminate ECM nonproducers in B. velezensis SQR9 biofilm, and try to determine the subpopulation for ECM (public goods) production and BAs (autotoxin) biosynthesis/BAs-induced self-immunity, as well as their interactions. We fused promoters for genes related to extracellular polysaccharides (EPS) and TasA fibers (two dominant ECM components in Bacillus biofilm Vlamakis et al., 2013) biosynthesis with mCherry, while the promoters for genes related to the autotoxin BAs biosynthesis and the self-immunity with gfp, obtained the transcriptional reporter Peps-mCherry, PtapA-mCherry, PbnaF-gfp, and PbnaAB-gfp, respectively. Their expression patterns were monitored using confocal laser scanning microscopy (CLSM) during the biofilm community formation. Photographs show that expression of the Peps-mCherry, PtapA-mCherry, PbnaF-gfp, and PbnaAB-gfp were all observed in a subpopulation cells of the whole community (Figure 1), which suggests a functional division of labor during biofilm formation; this cell differentiation pattern also indicates the ECM-nonproducers can be recognized as the cheater-like individuals (Otto et al., 2020). Importantly, the overlay of the double fluorescent reporters indicates that ECM and BAs production is generally raised in the same subpopulation (Figure 1; the yellow cells represent co-expression of mCherry and gfp), the flow cytometry also confirms the positive correlation between the two reporters within the picked cells as expected (Figure 1—figure supplement 1), since the self-immunity gene bnaAB was reported to be specifically activated by endogenous BAs (Huang et al., 2021), it was also preferentially expressed in the same subpopulation with ECM-producers (Figure 1, Figure 1—figure supplement 1). These results demonstrate general coordination of ECM production and BAs synthesis/immunity in the same subpopulation of the B. velezensis SQR9 biofilm community. Figure 1 with 1 supplement see all Download asset Open asset Expression of ECM production and BAs biosynthesis/immunity were located in the same subpopulation. Fluorescence emission patterns of double-labeled strains. Colony cells of different double-labeled strains were visualized using CLSM to monitor the distribution of fluorescence signals from different reporters. Peps-mCherry and PtapA-mCherry were used to indicate cells expressing extracellular polysaccharides (EPS) and TasA fibers production, respectively; PbnaF-gfp and PbnaAB-gfp were used to indicate cells expressing BAs synthesis and self-immunity, respectively. The bar represents 5 μm. Based on the co-expression pattern, we postulated that the ECM-nonproducing cheater-like cells, synchronously being sensitive to the BAs, could be killed by their siblings that produce both public goods ECM and the autotoxin BAs. Combining propidium iodide (a red-fluorescent dye for labeling dead cells) staining with reporter labeling, we monitored the cell death dynamics during the biofilm formation process in real-time. It was observed that a portion of the cells that didn’t produce public ECM or toxic BAs, or silenced in expression of the self-immunity gene bnaAB (cells without GFP signal), were killed by adjacent corresponding producers during the biofilm development process (Figure 2), while these producers remained alive throughout the incubation (Figure 2—videos 1–4); importantly, the number of dead cells adjacent to the producers was significantly higher than that closed to the non-producers (Figure 2—figure supplement 1). This lysis can be attributed to the BAs produced by the gfp-activated cells, as cannibalism of B. velezensis SQR9 was largely dependent on the production of this secondary metabolism (Huang et al., 2021). Taken together, the double-labeling observation and cell death dynamics detection indicate that the subpopulation of ECM and BAs producers selectively punish the nonproducing siblings, depending on a coordinately activated cell-differentiation pathway. Figure 2 with 5 supplements see all Download asset Open asset ECM and BAs producing subpopulations eliminated the nonproducing cheaters. The time-lapse experiment for observing the source and distribution of dead cells. Colony cells of different gfp-labeled strains were stained with propidium iodide (PI, a red-fluorescent dye for labeling dead cells) for 15 min, and then visualized by a CLSM to monitor the distribution of fluorescence signal from reporters and the PI dye. ‘0 min’ represents the time point at which cells are alive as shown by the arrow, ‘3 min’ or ‘6 min’ is the time point afterward, and the cells at the arrow die or even break apart. Peps-gfp and PtapA-gfp were used to indicate cells expressing extracellular polysaccharides (EPS) and TasA fibers production, respectively; PbnaF-gfp and PbnaAB-gfp were used to indicate cells expressing BAs synthesis and self-immunity, respectively. The total number of cells is 198 for strain SQR9-Peps-gfp, 71 for strain SQR9-PtasA-gfp, 88 for strain SQR9-PbnaF-gfp, and 162 for strain SQR9-PbnaAB-gfp. The bar represents 5 μm. Spo0A is the co-regulator for triggering ECM production and BAs synthesis/immunity To identify the potential co-regulator(s) of ECM production and BAs synthesis/immunity in B. velezensis SQR9, we evaluated the BAs production in an array of mutants that are known to be altered in ECM synthesis (ΔdegU, ΔcomPA, ΔabrB, ΔsinI, ΔsinR, and Δspo0A), by measuring their antagonism towards B. velezensis FZB42, a target strain specifically inhibited by BAs but no other antibiotics secreted by SQR9 (Wang et al., 2019). The BAs extract of wild-type SQR9 showed remarkable antagonism to the lawn of strain FZB42 (Figure 3A and B); only Δspo0A but no other mutants (all with the equal cell density of the wild-type), revealed a significantly reduced inhibition zone towards FZB42, and the complementary strain generally restored the antagonistic ability (Figure 3A and B). Spo0A is a well-investigated master regulator that governs multiple physiological behaviors in B. subtilis and closely-related species Hamon and Lazazzera, 2001; Molle et al., 2003; Xu et al., 2019b; as expected, the EPS production and biofilm formation was seriously impaired in Δspo0A (Figure 3—figure supplement 1). Intriguingly, Δspo0A but neither its complementary strain nor the wild-type, can be substantially inhibited by the BAs extract of strain SQR9, while Δspo0A was not inhibited by ΔGI3 that disabled in BAs production (Figure 3C), suggesting Spo0A does participate in the immunity to BAs. In addition, we constructed gfp transcriptional fusions to the promoter of genes involved in ECM production (eps & tapA) and BAs biosynthesis/immunity (bnaF/bnaAB) and discovered that under both liquid culture (Figure 3D) and plate colony conditions (Figure 3—figure supplement 2), their expression level was significantly decreased in Δspo0A as compared with the wild-type, which was restored in the complementary strain Δspo0A/spo0A. These results suggest that the global regulator Spo0A is the co-regulator for controlling ECM production and BAs biosynthesis/immunity in B. velezensis, which is probably dependent on the transcriptional regulation of certain relevant genes. Figure 3 with 2 supplements see all Download asset Open asset Spo0A is the co-regulator for triggering ECM production and BAs synthesis/immunity. (A) Oxford cup assay. Inhibition of the lawn of B. velezensis FZB42 by the BAs extract of wild-type SQR9, its different mutants altered in ECM production, and complementary strain Δspo0A/spo0A. (B) Quantification of inhibition zone. Diameter of the inhibition zones is observed in (A). (C) Oxford cup assay. Sensitivity of wild-type SQR9, Δspo0A, and Δspo0A/spo0A (as the lawn) to the extracellular extract of SQR9 and its mutant ΔGI3 that disable BAs synthesis. (D) Quantification of fluorescence in liquid culture. The expression level of eps, tapA, bnaF, and bnaAB in wild-type SQR9, Δspo0A, and Δspo0A/spo0A, as monitored by using gfp reporters fused to the corresponding promoters. Data are means and standard deviations from three biological replicates. * indicates a significant difference with the Control (SQR9) column as analyzed by Student’s t-test (p<0.05). Figure 3—source data 1 Related to Figure 3B. https://cdn.elifesciences.org/articles/84743/elife-84743-fig3-data1-v2.xlsx Download elife-84743-fig3-data1-v2.xlsx Figure 3—source data 2 Related to Figure 3D. https://cdn.elifesciences.org/articles/84743/elife-84743-fig3-data2-v2.xlsx Download elife-84743-fig3-data2-v2.xlsx Spo0A activates acetyl-CoA carboxylase (ACC) to support BAs synthesis and self-immunity In Bacillus, Spo0A governs the regulatory pathway for matrix gene (the eps and tapA-sipW-tasA operons) expression by controlling the activity of the regulators SinR and AbrB (Vlamakis et al., 2013), but how it mediates BAs synthesis and self-immunity remains unknown. We used biolayer interferometry analysis (BLI) for detecting molecular interaction signals between protein and DNA fragments (an increased signal during association and a decreased signal during dissociation). Results showed that the purified protein Spo0A cannot directly bind to the promoter of bnaF, suggesting it doesn’t induce BAs production through direct transcriptional activation (Figure 4—figure supplement 1). Alternatively, Spo0A has been reported to stimulate the expression of accDA that encodes ACC (Diomandé et al., 2015; Pedrido et al., 2013), which catalyzes acetyl-CoA to generate malonyl-CoA, an essential precursor for BAs biosynthesis (Figure 4A; Wang et al., 2019); therefore, we postulated accDA may be involved in the regulation of BAs production/immunity by Spo0A. We firstly verified the positive regulation of Spo0A on accDA expression in B. velezensis SQR9 by gfp fusion (Figure 4B, Figure 4—figure supplement 2 ). Since knockout of accDA, the essential gene for fatty acids biosynthesis, significantly impacts bacterial growth, we alternatively constructed a strain in which the original promoter of accDA was replaced by a xylose-inducible promoter (Pxyl), and monitored its BAs synthesis/immunity under different xylose induction conditions. The SQR9-Pxyl-accDA lost the antagonism ability towards target strain FZB42 in the absence of xylose, while the inhibition was significantly enhanced with the induction of xylose in a dose-dependent manner (Figure 4C and D). Since exogenous xylose didn’t influence the suppression of wild-type SQR9 on FZB42 (Figure 4C and D), these results suggest that accDA expression positively contributes to BAs production. Importantly, the SQR9-Pxyl-accDA was proved to be sensitive to SQR9-produced BAs without xylose addition, and the immunity was gradually restored with xylose supplement (Figure 4E, Figure 4—figure supplement 3). The xylose-induced transcription of accDA, also resulted in enhanced expression of genes involved in self-immunity (bnaAB; Figure 4F, for quantitative intensity please see Figure 4F, Figure 4—figure supplement 4A, C), but not BAs synthesis (bnaF; Figure 4F and Figure 4—figure supplement 4B, D), as the AccDA-derived malonyl-CoA accumulation affects BAs production in a post-transcriptional manner. The CLSM photographs and flow cytometry analysis also reveal that the activation of accDA (mCherry fusion) and bnaAB (gfp fusion) was located in the same subpopulation cells (Figure 4—figure supplement 5). Accordingly, these results indicate the positive regulation of Spo0A on BAs production/immunity in B. velezensis SQR9, is strongly dependent on accDA that encodes ACC. Figure 4 with 5 supplements see all Download asset Open asset Spo0A activates ACC for BAs synthesis and self-immunity. (A) Involvement of ACC in the biosynthesis of BAs in B. velezensis SQR9. ACC catalyzes acetyl-CoA to generate malonyl-CoA, which is transformed to malonyl-ACP under the catalyzation of ACP transacylase; then malonyl-ACP and acetyl-CoA are aggregated into a C5 primer, the precursor for BAs synthesis. (B) Quantification of fluorescence in liquid culture. The expression level of accDA in wild-type SQR9, Δspo0A, and Δspo0A/spo0A, as monitored by using the PaccDA-gfp reporter. (C) Oxford cup assay. Inhibition of the lawn of B. velezensis FZB42 by the BAs extract of wild-type SQR9 and SQR9-Pxyl-accDA, with the addition of different concentrations of xylose (0%, 0.1%, and 0.2%). (D) Quantification of inhibition zone. Diameter of the inhibition zones is observed in (C). (E) Oxford cup assay. Sensitivity of wild-type SQR9 and SQR9-Pxyl-accDA (as the lawn) to the BAs extract of SQR9 (100 μL (1x) or 200 μL (2x)), with the addition of different concentrations of xylose (0%, 0.1%, and 0.2%). (F) Colony fluorescence. Expression of bnaF and bnaAB in the colony cells of wild-type SQR9 and SQR9-Pxyl-accDA, with the addition of different concentrations of xylose (0%, 0.1%, and 0.2%). Colonies were observed under both bright fields (BF in the figure) and GFP channel, to monitor the fluorescence of PbnaF-gfp and PbnaAB-gfp reporters in different strains. The bar represents 1 mm. Data are means and standard deviations from three biological replicates. * in (B) indicates a significant difference (p<0.05) with the Control (SQR9) column as analyzed by Student’s t-test; columns with different letters in (D) are statistically different according to Duncan’s multiple range test (‘a’ for wild-type SQR9 under different concentrations of xylose and ‘a'’ for SQR9-Pxyl-accDA; p<0.05). Figure 4—source data 1 Related to Figure 4B. https://cdn.elifesciences.org/articles/84743/elife-84743-fig4-data1-v2.xlsx Download elife-84743-fig4-data1-v2.xlsx Figure 4—source data 2 Related to Figure 4D. https://cdn.elifesciences.org/articles/84743/elife-84743-fig4-data2-v2.xlsx Download elife-84743-fig4-data2-v2.xlsx The co-regulation policing system optimizes the division of labor and promotes population fitness Having illustrated the molecular mechanism of the co-regulation pathway for punishing nonproducing cheater-like cells in B. velezensis SQR9, we wondered about the broad-spectrum ecological significance of this policing system for B. velezensis SQR9 at a community level. We constructed two mutants with disabled sanction mechanism, the ΔbnaV deficient in BAs synthesis (loss of the punishing weapon) and the SQR9-P43-bnaAB that continually expresses the self-immunity genes (cheater-like individuals cannot be punished by the weapon BAs), both mutants showed similar growth characteristics with the wild-type (Figure 5—figure supplement 1). We first applied flow cytometry analysis to test whether the lack of the policing system (ΔbnaV and SQR9-P43-bnaAB) impairs the punishment of public goods-nonproducers during biofilm formation. The proportion of matrix-producing cooperators (eps & tapA active cells) in the wild-type community, as well as the average expression level of corresponding genes, were significantly higher than that in the ΔbnaV or SQR9-P43-bnaAB community (Figure 5A and B), suggesting the division of labor in the two mutants population was significantly different with the wild-type. Consequently, the wild-type established a more vigorous biofilm as compared with the two mutants, as shown by the earlier initial progress, larger maximum biomass, and delayed dispersal process (prolonged stationary phase) (Figure 5C and D). Additionally, the robust biofilm formed by the wild-type also endowed them with stronger resistance against different stresses, including antibiotics, salinity, acid-base, and oxidation (Figure 5D, Figure 5—figure supplement 2 and Figure 5—figure supplement 3). These data indicate the policing system in wild-type SQR9 ameliorates the division of labor during biofilm formation, thereby promoting community fitness. Figure 5 with 4 supplements see all Download asset Open asset The co-regulation policing system optimizes the division of labor and enhances population fitness. (A) Flow cytometry monitoring the expression of Peps-gfp and PtapA-gfp reporters in wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB. (B) Quantification of (A). The proportion of the active cells (%) and average FITC in wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB, as monitored by Peps-gfp and PtapA-gfp reporters using flow cytometry. (C) Pellicle morphology. Pellicle formation dynamics of wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB in MSgg medium. (D) Quantification of pellicles. Pellicle weight dynamics of wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB in MSgg medium under normal (corresponds to (C)) or stressed conditions (H2O2, tetracycline, or 7% NaCl). (E) Qualitative analysis of protease or siderophore yield. Production of proteases and siderophore by wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB colonies. (F) Root colonization assay. Comparison of root colonization of wild-type SQR9, SQR9ΔbnaV, and SQR9-P43-bnaAB. Data are means and standard deviations from three biological replicates; columns with different letters are significantly different according to Duncan’s multiple range tests, p<0.05. Figure 5—source data 1 Related to Figure 5B. https://cdn.elifesciences.org/articles/84743/elife-84743-fig5-data1-v2.xlsx Download elife-84743-fig5-data1-v2.xlsx Figure 5—source data 2 Related to Figure 5D. https://cdn.elifesciences.org/articles/84743/elife-84743-fig5-data2-v2.xlsx Download elife-84743-fig5-data2-v2.xlsx Figure 5—source data 3 Related to Figure 5F. https://cdn.elifesciences.org/articles/84743/elife-84743-fig5-data3-v2.xlsx Download elife-84743-fig5-data3-v2.xlsx Besides the well-known regulation of biofilm matrix production, Spo0A also controls the production of other public goods such as proteases and siderophore (Fujita et al., 2005; Molle et al., 2003); it can be recognized as a critical switch that governs the cell transition from a free-living and fast-growing status (Spo0A-OFF), to a multicellular and cooperative style (Spo0A-ON) (Shank and Kolter, 2011; López et al., 2009c). Intrinsically, the punishing targets of this policing system are supposed not limited to the cheater-like matrix-nonproducers, but all of the Spo0A-OFF individuals (cells that don’t express the immune genes bnaAB, including protease-nonproducers and siderophore-nonproducers). Therefore, we determined the production of extracellular proteases and siderophore among the three strains, revealing that these public goods were also accumulated more in the wild-type than in these two mutants’ communities (Figure 5E, Figure 5—figure supplement 4). Importantly, the wild-type SQR9 demonstrated significantly stronger root colonization compared with the two mutant strains losing the cheater punishing system (Figure 5F). In summary, the Spo0A governed co-regulation punishment system effectively optimizes the division of labor and altruistic behavior in the B. velezensis population, by excluding the cheater-like nonproducers to a certain degree, consequently improving the population stability and ecological fitness under different conditions. Discussion Division of labor, where subpopulations perform complementary tasks simultaneously within an assembly, characterizes major evolutionary transitions of cooperation in certain cases (Babak, 2018). Unlike the diverse strategies for preventing obligate cheaters in cooperative systems (Özkaya et al., 2017; Smith and Schuster, 2019; Travisano and Velicer, 2004), division of labor requires an efficiency benefit and alignment of interests covering different specialized individuals (West and Cooper, 2016). For instance, compared with cells that produce a certain kind of public goods (e.g. ECM or extracellular hydrolases), the subpopulations that don’t perform these tasks (but still share these benefits) become cheater-like individuals, and their proportion needs to be controlled for maintaining community stability and fitness (Martin et al., 2020; West and Cooper, 2016). In the present study, we demonstrated that during biofilm formation, the beneficial rhizobacterium B. velezensis SQR9 engages a policing system that coordinately actives ECM production and autotoxin synthesis/immunity, to punish the cheater-like subpopulation silencing in public goods secretion and restrain their pro