ABSTRACT
Efficient biofilm formation and root colonization capabilities facilitate the ability of beneficial plant rhizobacteria to promote plant growth and antagonize soilborne pathogens. Biofilm formation by plant-beneficial Bacillus strains is triggered by environmental cues, including oxygen deficiency, but the pathways that sense these environmental signals and regulate biofilm formation have not been thoroughly elucidated. In this study, we showed that the ResDE two-component regulatory system in the plant growth-promoting rhizobacterium Bacillus amyloliquefaciens strain SQR9 senses the oxygen deficiency signal and regulates biofilm formation. ResE is activated by sensing the oxygen limitation-induced reduction of the NAD+/NADH pool through its PAS domain, stimulating its kinase activity, and resulting in the transfer of a phosphoryl group to ResD. The phosphorylated ResD directly binds to the promoter regions of the qoxABCD and ctaCDEF operons to improve the biosynthesis of terminal oxidases, which can interact with KinB to activate biofilm formation. These results not only revealed the novel regulatory function of the ResDE two-component system but also contributed to the understanding of the complicated regulatory network governing Bacillus biofilm formation. This research may help to enhance the root colonization and the plant-beneficial efficiency of SQR9 and other Bacillus rhizobacteria used in agriculture.
IMPORTANCE Bacillus spp. are widely used as bioinoculants for plant growth promotion and disease suppression. The exertion of their plant-beneficial functions is largely dependent on their root colonization, which is closely related to their biofilm formation capabilities. On the other hand, Bacillus is the model bacterium for biofilm study, and the process and molecular network of biofilm formation are well characterized (B. Mielich-Süss and D. Lopez, Environ Microbiol 17:555–565, 2015, https://doi.org/10.1111/1462-2920.12527; L. S. Cairns, L. Hobley, and N. R. Stanley-Wall, Mol Microbiol 93:587–598, 2014, https://doi.org/10.1111/mmi.12697; H. Vlamakis, C. Aguilar, R. Losick, and R. Kolter, Genes Dev 22:945–953, 2008, https://doi.org/10.1101/gad.1645008; S. S. Branda, A. Vik, L. Friedman, and R. Kolter, Trends Microbiol 13:20–26, 2005, https://doi.org/10.1016/j.tim.2004.11.006; C. Aguilar, H. Vlamakis, R. Losick, and R. Kolter, Curr Opin Microbiol 10:638–643, 2007, https://doi.org/10.1016/j.mib.2007.09.006; S. S. Branda, J. E. González-Pastor, S. Ben-Yehuda, R. Losick, and R. Kolter, Proc Natl Acad Sci U S A 98:11621–11626, 2001, https://doi.org/10.1073/pnas.191384198). However, the identification and sensing of environmental signals triggering Bacillus biofilm formation need further research. Here, we report that the oxygen deficiency signal inducing Bacillus biofilm formation is sensed by the ResDE two-component regulatory system. Our results not only revealed the novel regulatory function of the ResDE two-component regulatory system but also identified the sensing system of a biofilm-triggering signal. This knowledge can help to enhance the biofilm formation and root colonization of plant-beneficial Bacillus strains and also provide new insights of bacterial biofilm formation regulation.
INTRODUCTION
Plant growth-promoting rhizobacteria (PGPR), which have the remarkable ability to suppress soilborne pathogens and stimulate plant growth, have been widely used in agriculture (1, 2). It is known that efficient root colonization of PGPR is necessary for them to exert their beneficial effects in the rhizosphere, and root colonization is related to the ability of PGPR to form biofilms (3, 4). Biofilms are multicellular communities that are held together by self-produced extracellular matrices, enabling bacteria to survive in unsuitable and resource-limited environments (5, 6). As a model bacterium, the characteristics and molecular mechanisms of Bacillus subtilis biofilm formation have been deeply investigated (7). This Gram-positive bacterium forms complex biofilm structures at air-liquid interfaces (pellicles) and on solid surfaces (colonies) (7, 8). The extracellular matrix is the predominant structural component of biofilms, consisting of exopolysaccharides (EPS) encoded by the epsA-O operon (9), TasA protein fibers encoded by the tapA-sipW-tasA operon (10–12), the hydrophobic small secreted protein BslA (13, 14), and extracellular DNA (eDNA) (15). Fragile pellicles and smooth-textured colonies are formed when cells are unable to synthesize EPS or TasA properly.
The synthesis of the extracellular matrix for biofilm formation is regulated by a complicated molecular network. The phosphorylation of Spo0A, which is mediated by five histidine kinases (KinA to KinE), is a critical regulation process (16–18). SinR and AbrB act as negative regulators in the biofilm formation process by suppressing the expression of the matrix genes (13, 19, 20). Phosphorylated Spo0A (Spo0A∼P) represses the expression of AbrB to release its suppression for extracellular matrix production. Spo0A∼P can also release the SinR-mediated suppression by activating the expression of a small protein, SinI, which is an antirepressor of SinR (17, 21). The relief of the repression caused by SinR also activates the synthesis of SlrR, a helix-turn-helix-containing protein that is required for full expression of the tapA operon (22, 23).
Biofilm formation by B. subtilis is triggered by environmental cues (24). Previous studies have demonstrated that tomato roots secrete l-malic acid, and a combination of glycerol and manganese could trigger biofilm formation in a manner that is dependent on the KinD CACHE domain (25, 26). ClO2 activates the KinC kinase by causing a collapse of the membrane potential (26, 27). Oxygen limitation induces biofilm formation through impaired respiration, which activates Spo0A phosphorylation via KinB. KinA also contributes to the impaired respiration response by sensing a change in the NAD+/NADH pool in a PAS domain-dependent manner (28). Some Bacillus strains secrete small molecules to stimulate their biofilm formation, and the surfactin produced by B. subtilis stimulates biofilm formation through the potassium leakage-mediated activation of the membrane protein kinase KinC (29, 30). An antifungal lipopeptide (bacillomycin D) secreted by Bacillus amyloliquefaciens strain SQR9 was reported to contribute to its biofilm formation (4).
Two-component regulatory systems are suggested to be primarily responsible for recognizing environmental signals and initiating cellular signaling circuits, facilitating rapid and precise responses to environmental alterations (31, 32). Twenty-nine two-component regulatory systems (including DegSU, ComPA, and LytST) in B. subtilis have been reported to play key roles in physiological processes (32). Increased DegU phosphorylation significantly improved the complex colony architecture, biofilm formation, root colonization, and antibiotic production of B. amyloliquefaciens SQR9 (33). In B. subtilis, the ComPA system functions in the development of competence, the production of the lipopeptide antibiotic surfactin, degradative enzyme production, and even some unknown functions (34, 35). The LytST system is believed to function in programmed cell death (PCD) and autolysis by sensing a decrease in membrane potential, while genomic DNA released by the lysed bacteria is essential for biofilm stability and intercellular adhesion (36–38).
Our previous work showed that the ResDE two-component regulatory system regulates biofilm formation in B. amyloliquefaciens SQR9 (39), but the triggering environmental signal and the downstream molecular pathway remained unclear. Many studies have revealed the regulatory role of ResDE in both aerobic and anaerobic respiration in bacteria (40–44). Upon sensing a decrease in oxygen pressure, ResE activity changes from that of a phosphatase to a kinase, and it then transfers a phosphoryl group to the response regulator ResD; ResD∼P improves the transcription of targeted genes involved in the synthesis of heme-containing components of the electron transport chain (45). Another study reported on the oxygen limitation-induced biofilm formation in B. subtilis (28), but how the signal of decreased oxygen pressure is recognized is unknown. On the basis of our previous work on the function of the ResDE two-component regulatory system with respect to biofilm formation (39) and its well-studied regulatory role in respiration, we hypothesized that the ResDE system of B. amyloliquefaciens SQR9, which was reassigned as “operational group B. amyloliquefaciens” according to Fan et al. (46), senses the induced oxygen pressure signal and regulates biofilm formation in strain SQR9.
In this study, we provide evidence showing that ResE is activated by sensing a decrease in the NAD+/NADH ratio caused by oxygen limitation in a PAS domain-dependent manner. This stimulates its kinase activity, after which, it transfers the phosphoryl group to ResD. The phosphorylated ResD directly binds to the promoter regions of the qoxABCD and ctaCDEF operons to improve the biosynthesis of terminal oxidases, which can interact with KinB to activate biofilm formation. Our results have elucidated the specific mechanism by which the ResDE system regulates biofilm formation, especially with respect to the sensing of environmental oxygen limitation by ResDE that leads to biofilm formation. The results of our study provide deep insights into the complicated regulatory network of Bacillus biofilm formation.
RESULTS
ResDE two-component regulatory system senses a reduced oxygen signal to induce biofilm formation in B. amyloliquefaciens SQR9.To test our hypothesis that the ResDE two-component regulatory system of B. amyloliquefaciens SQR9 senses a decreased oxygen pressure signal and regulates biofilm formation in strain SQR9, we first investigated the effect of oxygen on the biofilm formation of strain SQR9. The results showed that as the oxygen concentration decreased from 22% to 5 to 8%, wild-type SQR9 colonies became more intensely wrinkled (Fig. 1A). In contrast, increasing the oxygen concentration from 22% to 35% suppressed the formation of wrinkles, resulting in flat and smooth colonies (Fig. 1A). These results suggested that decreased oxygen is a signal that elicits biofilm formation. Oxygen deficiency increased the expression of tapA and epsD, genes important in the production of the SQR9 biofilm matrix substance, suggesting that the oxygen deficiency-induced biofilm formation of strain SQR9 depends on increased extracellular matrix production (Fig. 1B).
Colony structures of Bacillus amyloliquefaciens SQR9, ΔresD, ΔresE strains under different oxygen concentrations and transcriptional analysis of extracellular matrix genes and resD-E under a limited oxygen condition. (A) Colony wrinkling increased with decreasing oxygen concentration. A high oxygen concentration suppressed wrinkling, while the responses of colonies to the different oxygen concentrations were blocked with deletion of resD or resE. Transcriptional levels of epsD and tapA (B) and resD-E (C) were upregulated under oxygen-limiting conditions relative to that under the normal oxygen concentration. The recA gene was used as an internal reference gene. Bars represent the standard deviations of data from three biological replicates. (D) PresD-gfp showed intense fluorescence under oxygen-limiting conditions. (E) NAD+/NADH ratios for the wild-type SQR9 strain grown for 48 h on solid biofilm-inducing medium under the indicated oxygen concentrations. The NAD+/NADH ratios are the averages from six colonies for each oxygen concentration.
As previously reported, SQR9 mutants of ΔresE and ΔresD cannot form normal pellicles of biofilm (39). Here, we proved that these strains also lose the ability to respond to oxygen deficiency and form complex structured colonies (Fig. 1A), though the growth curves of the strains were similar (see Fig. S1 in the supplemental material). This indicated that the ResDE two-component regulatory system might mediate oxygen deficiency-induced biofilm formation. To test this hypothesis, the expression of resD and resE was measured under oxygen-limiting conditions using quantitative PCR (qPCR) and was compared with the expression of these genes under a normal oxygen concentration. The results showed that decreased oxygen concentrations increased the expression of resD and resE (Fig. 1C). When the resD promoter was fused with a gfp reporter gene (PresD-gfp), the reporter strain only showed weak fluorescence under the normal oxygen condition, whereas the fluorescence intensity was much stronger under an oxygen concentration of 5 to 8% (Fig. 1D). These results suggested that the transcription of the ResDE two-component system could be induced by the signal of oxygen limitation.
To explore how ResE senses the oxygen limitation signal, the subdomains of the ResE protein were predicted, which showed that the signal input domain of ResE consists of two transmembrane regions, a HAMP linker and a PAS domain (see Fig. S2). The PAS domain is primarily responsible for sensing environmental signals, including oxygen, light, redox state, and cellular energy levels (47). We speculated that the imbalance of the cellular redox state (NAD+/NADH) caused by oxygen deficiency is sensed by ResE. The NAD+/NADH ratio decreased from ∼1.2 to ∼0.4 when the oxygen concentration decreased from 22% to 5 to 8% (Fig. 1E). The results of the isothermal titration calorimetry (ITC) assays indicated that purified ResE specifically binds to NAD+ (Fig. 2A), but that binding did not occur between ResE and NADH (Fig. 2B). On the basis of these results, we concluded that ResE senses the decrease in the NAD+/NADH ratio caused by oxygen limitation and then activates ResD via phosphorylation.
Isothermal titration calorimetry assay of ResE-NAD+ and ResE-NADH interactions. The upper panels show the heat changes observed upon the addition of 40 μl of a 1.05 mM NAD+ (A) or NADH (B) solution in PBS buffer (pH 7.4) into a 70 μM ResE protein solution in the same buffer. The lower panels show the integrated heat changes of each injection plotted against the molar ratio of NAD+ and NADH to the ResE protein.
Furthermore, mutations in resE and resD decreased the expression of the biofilm extracellular matrix genes tapA and epsD, the global regulator spo0A, and the antirepressor sinI, while the global repressor abrB was upregulated (Fig. 3A). These results further suggested that the ResDE system is involved in oxygen deficiency-induced biofilm formation.
Transcriptional levels of biofilm-related and cytochrome complex synthesis genes in ΔresD and ΔresE mutants. (A) The transcriptional levels of spo0A, abrB, sinI, epsD, and tapA in res mutants relative to that of the wild-type strain evaluated by qPCR. Transcriptional levels of ctaCDEF and ctaB (B), qcrABC (C), and qoxABCD (D) operons in res mutants relative to those of the wild-type strain evaluated by qPCR. The recA gene was used as an internal reference gene. Bars represent standard deviations of data from three biological replicates. RQ, relative quantification.
ResDE participates in the biofilm formation process via the regulation of the terminal oxidases of the electron transport chain.Kolodkin-Gal et al. reported that oxygen deficiency induced B. subtilis strain NCIB3610 biofilm formation because of the impaired respiration (28). In this study, we hypothesized that the ResDE two-component system senses the oxygen deficiency signal and regulates the terminal oxidases of the electron transport chain to affect biofilm formation. Oxygen is the electron acceptor of respiration. When the biofilm-inducing medium was supplemented with 20 mM KNO3, the formation of wrinkles on the colonies was suppressed, even under an oxygen concentration of 5 to 8% (see Fig. S3). This suggested that the nitrate could serve as an alternate electron acceptor to restore the impaired respiration, thus eliminating the response of strain SQR9 to a low oxygen concentration. These results indicated that the oxygen deficiency-induced SQR9 biofilm formation was via impaired respiration. The expression of the ctaCDEF operon and ctaB (encoding cytochrome caa3), the qcrABC operon (encoding cytochrome bc), and the qoxABCD operon (encoding cytochrome aa3) was downregulated in the res mutant strains compared with that in the SQR9 wild-type strain, which suggested that the ResDE system affects the expression of the ctaCDEF, qcrABC, and qoxABCD operons and ctaB (Fig. 3B to D).
To determine whether ResD regulates these operons by direct binding to their promoter regions, electrophoretic mobility shift assays (EMSAs) and DNase I footprint assays were performed using purified ResD. The results of the EMSAs showed that ResD retarded the electrophoretic mobility of a DNA fragment containing the Pqox and Pcta regions in a concentration-dependent manner (Fig. 4A and B), indicating that the ResD protein can directly and specifically bind to the promoter regions of qoxABCD and ctaCDEF operons. To define the ResD-binding site in the promoter regions of the qoxABCD and ctaCDEF operons, dye-based DNase I footprint assays were performed by using dye primer sequencing. These assays pinpointed the direct repeat (DR) motifs (TTTTGACGATTTTG, from +105 to +119 and TTTTATGAATTTTA, from −55 to −41) in the middles of these protected regions of the qoxABCD and ctaCDEF operons (Fig. 4C and D) (48). However, no binding was observed between ResD and the qcrABC operon promoter (data not shown). These results suggested that ResD could directly regulate the synthesis of cytochromes caa3 and aa3 (encoded by the ctaCDEF and qoxABCD operons, respectively), while its effect on cytochrome bc (encoded by the qcrABC operon) appears to be indirect.
Electrophoretic mobility shift assays (EMSAs) and DNase I footprint assays of ResD-ctaCDEF operon and ResD-qoxABCD operon promoters' interactions. (A, B) EMSAs in which end-labeled DNAs (indicated at the bottom of each figure) were mixed with purified ResD at the following concentrations: 0, 0.28, 0.7, 1.4, and 2.1 μM. (C, D) Electropherograms showing the protected regions of the qoxABCD and ctaCDEF operon promoters after digestion with DNase I following an incubation in the absence or the presence of 3.2 μg and 0.8 μg ResD protein, respectively.
Terminal oxidases contribute to biofilm formation by interacting with KinB.To investigate the effects of the cytochrome components on the biofilm formation of strain SQR9, Δcta, Δqcr, and Δqox mutants were constructed and their biofilm formation abilities were evaluated under various oxygen concentrations. The Δqox strain lost the ability to enhance biofilm formation in response to decreased oxygen concentrations, although the growth curve was similar to that of the wild-type strain, but the Δcta and Δqcr mutants showed colony morphologies similar to that of the wild-type SQR9 strain (Fig. 5A; Fig. S1). The Δqox mutant also showed decreased transcription of the biofilm extracellular matrix genes epsD and tapA, the master regulator spo0A, the antirepressor sinI, and the kinase kinB compared with that of the wild-type SQR9 strain, whereas the global repressor abrB was upregulated (Fig. 5B). These results suggested that the disruption of the qoxABCD operon impaired the process of biofilm formation by disrupting extracellular matrix synthesis.
Colony structures of Δcta, Δqcr, and Δqox mutants under different oxygen concentrations and the transcriptional and BTH assays. (A) A lack of the qoxABCD operon blocks the response of colonies to different oxygen concentrations, while a deficiency of either the ctaCDEF or qcrABC operon does not. (B) The transcriptional levels of spo0A, abrB, sinI, kinB, epsD, and tapA in a Δqox mutant relative to that in the wild-type strain evaluated by qPCR. The recA gene was used as an internal reference gene. Bars represent standard deviations of data from three biological replicates. (C) BTH analysis to study the interaction between cytochrome aa3 and KinB. The positive interaction between QoxABCD and KinB is shown by the blue colonies and the quantification of the β-galactosidase activity (1,400 to 1,500 Miller units). The strains harboring plasmids (pKT25-zip plus pUT18C-zip) and empty plasmids (pKNT25 plus pUT18) are the positive and negative controls, respectively. The LytT strain was also used as a negative control. Dashed line indicates the threshold limit that defines a positive (≥700 Miller units) or a negative (<700 Miller units) interaction signal according to the manufacturer's instructions (EuroMedex). The results represent the means from three independent experiments. Asterisks denote significant differences according to Duncan's multiple range tests at a P value of <0.05.
The impaired expression of kinB and spo0A in the Δqox mutant suggested that cytochrome aa3 likely interacts with KinB and then activates the biofilm formation process. The deletion of kinB impaired the ability to respond to the limited oxygen concentration, though the growth curve was similar to that of the wild-type strain (see Fig. S1 and S4). Bacterial two-hybrid (BTH) assay results suggested that a positive interaction indeed exists between the four subunits of cytochrome aa3 (QoxA, QoxB, QoxC, and QoxD) and KinB (Fig. 5C). On the basis of these results, we showed that the kinase KinB was involved in the ResDE regulatory circuit for biofilm formation in response to oxygen deficiency via a protein-protein interaction.
DISCUSSION
We previously reported that a novel regulatory protein (the two-component regulatory system protein ResE) governs biofilm formation in B. amyloliquefaciens SQR9 (39). In this study, we attempted to elucidate the detailed regulatory mechanism of ResDE with respect to biofilm formation. The results showed that decreased oxygen pressure serves as an upstream signal to activate the ResDE system, after which the phosphorylated ResD protein activates the expression of terminal oxidases by directly binding to the promoter regions of their encoding genes. Finally, the terminal oxidases induced biofilm formation in a protein-protein interaction-dependent manner with KinB, which regulates the main Spo0A pathway involved in Bacillus biofilm formation (Fig. 6).
Model for the two-component regulatory system ResDE mediation of biofilm formation by the control of the electron transport chain under oxygen-limiting conditions (see description in the text).
Generating timely responses to complex and sophisticated environmental signals is necessary for microbes to survive in natural niches, with the key process of the initial cue sensing and signal transduction performed primarily by two-component regulatory systems. During the regulatory process, cross-regulation can provide a means of integrating different environmental signals into a harmonized output response (49). The drop in the NAD+/NADH ratio was observed to be sensed by ResE in our study of B. amyloliquefaciens SQR9 and was sensed by KinA in a PAS domain-dependent manner in Kolodkin-Gal's study of B. subtilis NCIB3610 (28), suggesting that distinct pathways are applied to integrate environmental signals and facilitate rapid and precise responses to environmental turbulence.
B. subtilis and B. amyloliquefaciens respond to impaired respiration via interactions of the histidine kinase KinB with the aerobic respiratory apparatus, which is more like a compensatory mechanism to help cells in the colonies to survive under oxygen-deficient conditions (28). In particular, the stimulation of KinB activity is controlled by a functional (but not fully functioning) cytochrome apparatus (28). In this study, we provide evidence that a lack of ResD/E affects the transcriptional levels of the cytochrome apparatus, including ctaCDEF, qcrABC, and qoxABCD (Fig. 3B to D), and that ResD can directly bind to the qoxABCD and ctaCDEF operons (Fig. 4A and B). These results suggest that ResDE plays a key role in connecting environmental signals and multicellular behavior.
Although both oxygen deficiency and high iron concentrations induced the intense wrinkling of colonies, the specific mechanisms of the two environmental cues appear to be different (28). Limited oxygen levels appear to be a stimulus that activates ResDE to promote the biosynthesis of the cytochrome apparatus and improve aerobic respiration efficiency, an optimal strategy for cells to adapt to oxygen-deficient conditions. In contrast, a high iron concentration affects respiration efficiency by improving the assembly of the cytochromes in the respiratory apparatus. In this study, the Δqox mutant lost the ability to respond to variations in oxygen levels, forming flat colonies that were similar to those produced by the ΔresE and ΔresD strains (Fig. 5A). The phenotypes of Δcta and Δqcr mutants were similar to that of the wild-type SQR9 strain (Fig. 5A), suggesting a key role of cytochrome aa3 in the response of Bacillus cells to oxygen deficiency. In a work by Kolodkin-Gal et al., the expression of sinI was decreased under normal or high iron concentrations in a mutant lacking ctaCDEF and qcrABC (28). It is reasonable to suggest that both cytochrome aa3 and cytochrome caa3 play essential roles in the biofilm formation process but do so in different manners. Additionally, Stanley et al. found by using DNA microarrays that the transcriptional levels of the resDE, ctaCDEF, qoxABCD, and qcrABC operons were upregulated when the cells transformed from a planktonic to a biofilm state, suggesting that all these operons could be involved in the biofilm formation process (50).
Despite the process of biofilm formation having been well studied, the intricate networks governing biofilm formation still need to be explored. Our study on how the ResDE two-component regulatory system participates in the biofilm formation process in B. amyloliquefaciens SQR9 suggests that bacteria might exploit multiple approaches to rapidly adapt a multicellular behavior to complex environments.
MATERIALS AND METHODS
Strains and culture conditions.The strains and plasmids used in this study are described in Table 1. Bacillus amyloliquefaciens SQR9 (CGMCC accession no. 5808; China General Microbiology Culture Collection Center) was used in this study. Escherichia coli TOP10 was used as the host strain for all plasmids. E. coli BL21 was used for the heterologous expression of targeted proteins. Strain SQR9 was cultured in Luria-Bertani (LB) medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl) at 37°C. Antibiotics were added as required at the following concentrations: 20 μg/ml Zeocin, 1 μg/ml erythromycin (Em), and 5 μg/ml of chloramphenicol. E. coli TOP10 cells were grown in LB medium at 37°C. When necessary, 30 μg/ml of kanamycin or 12.5 μg/ml chloramphenicol was added. For pellicle formation studies, MSgg medium (5 mM potassium phosphate [pH 7.0], 100 mM MOPS [morpholinepropanesulfonic acid] [pH 7], 2 mM MgCl2, 700 mM CaCl2, 50 μM MnCl2, 50 μM FeCl3, 1 μM ZnCl2, 2 μM thiamine, 0.5% glycerol, 0.5% glutamate, 50 μg/ml tryptophan, 50 μg/ml phenylalanine) was used (7).
Strains and plasmids used in this study
B. amyloliquefaciens SQR9 mutant construction.To delete the targeted genes in B. amyloliquefaciens SQR9, 1-kb fragments located upstream and downstream of the target gene were amplified. The 1.6-kb erythromycin resistance gene (Emr) was amplified from plasmid pAX01 (51), which partially overlapped the upstream and downstream fragments of each target gene. The upstream and downstream fragments were fused with the erythromycin resistance gene by overlap extension. Primers used in this experiment are shown in Table 2. The 25-μl mixture for the first step of the overlapping PCR contained 5 μl PrimeSTAR buffer (5×), 2 μl deoxynucleoside triphosphate (dNTP) mix, 1 μl of the upstream fragment, 1 μl of the downstream fragment, 1 μl of the Emr fragment, 0.3 μl PrimeSTAR HS DNA polymerase, and 14.7 μl double-distilled water (ddH2O); the PCR program used was as follows: 98°C for 2 min, 10 cycles of 98°C for 10 s, 52°C for 10 s, and 72°C for 4.5 min. The 25-μl mixture for the second step of the overlapping PCR contained 5 μl PrimeSTAR buffer (5×), 2 μl dNTP mix, 1 μl of the forward primer for the upstream fragment, 1 μl of the reverse primer for the downstream fragment, 1 μl of the PCR product from the first step, 0.3 μl PrimeSTAR HS DNA polymerase, and 14.7 μl ddH2O. The PCR program used was as follows: 98°C for 2 min, 32 cycles of 98°C for 10 s, 57°C for 10 s, and 72°C for 4.5 min. All PCR products were purified with an AxyPrep DNA gel purification and extraction kit (Axygen, Hangzhou, China).
Primers used in this study
The fused fragments were transformed into SQR9 competent cells (52), and the transformants were selected on LB agar plates containing 1 μg/ml Em (see Fig. S5 in the supplemental material). The mutants were verified by PCR using specific primer sets (Table 2).
B. amyloliquefaciens SQR9 biofilm architecture under a limited oxygen condition.An AnaeroPack (MGC, Japan) capable of reducing the oxygen concentration from 22% to 5 to 8% was used in this experiment. The mutant and wild-type SQR9 strains were incubated in LB medium until the optical density at 600 nm (OD600) reached approximately 1.0. The cells were collected, washed twice with 0.01 M phosphate-buffered saline (PBS), and suspended in the same buffer in an equivalent volume. The bacterial suspensions were aliquoted on the MSgg plates to form colonies, and the plates were incubated for 24 to 36 h at 37°C to observe the biofilm architectures. In addition, 20 mM KNO3 was added to the MSgg medium plates as an alternative electron acceptor.
Gene transcription analyses by qPCR.Pellicles that formed in the wells of the plate were collected. The extraction of RNA was performed according to instructions supplied by the E.Z.N.A. bacterial RNA kit (OMEGA, Biotek, USA). The reverse transcription of RNA was conducted with the PrimeScript RT reagent kit with a genomic DNA (gDNA) eraser (TaKaRa Biotechnology, Dalian). The genes involved in biofilm formation regulation and extracellular matrix synthesis (spo0A, sinI, abrB, epsD, tapA, and kinB) and components of the electron transport chain synthesis (ctaBCDEF, qcrABC, and qoxABCD) were quantified by qPCR, and the recA gene was used as an internal control (Table 2). The 2−ΔΔCT method was used to analyze the qPCR data (53).
Quantification of the NAD+/NADH ratio.The preparation of samples was carried out according to the method described by Kolodkin-Gal et al. (28). For the quantification of NAD+/NADH, the wild-type strain was incubated on solid MSgg medium for 48 h. Six colonies from each sample were harvested in 1 ml of PBS (pH 7.4) and were subjected to mild sonication to separate the extracellular matrices of cells. The cells were pelleted by centrifugation at 15,000 rpm for 1 min and washed with cold PBS buffer. Next, the cells were resuspended in 400 μl NAD+/NADH extraction buffer, and 30 μl of a 50 mg/ml lysozyme solution was added. To remove the lysozyme, samples were filtered with 10-kDa-molecular mass-cutoff filters. To decompose NAD+, 200-μl aliquots of extracted samples were heated to 60°C for 30 min. Next, the assay was performed according to the instructions of an NAD+/NADH quantification colorimetric kit (BioVision, USA).
Expression and purification of the ResD and ResE proteins.The ResD and ResE proteins were expressed in E. coli BL21 carrying pET29a and constructs and were purified with Ni-nitrilotriacetic acid (NTA) agarose (Thermo Fisher Scientific, USA) (54, 55). The plasmids (pET29a-resD and pET29a-resE) were constructed as follows. The resD coding fragment was amplified by PCR using SQR9 genomic DNA with the primer set E-resDPF/E-resDPR. A fragment coding for the ResE cytoplasmic subdomain (including the PAS, HisKA, and HATPase_C subdomains of ResE) was amplified by PCR using SQR9 genomic DNA with the primer set E-resEPF/E-resEPR. The amplified fragments were inserted in pET29a using the restriction sites XhoI and KpnI.
Isothermal titration calorimetry assay.For the ITC assays, measurements were performed on an ITC200 (MicroCal) at 25°C, and the data were extracted and processed using the Origin software package. The ResE protein was dialyzed against PBS buffer (pH 7.4). For ligand binding analyses, 70 μM protein was introduced into the sample cell and was titrated with aliquots of the ligand solution. A total of 20 injections, all 40 μl in volume, were performed.
Analysis of promoter regions.Fragments carrying the gfp gene sequence without the promoter region were amplified from plasmid pHAPII (56) and then were cloned into the B. amyloliquefaciens-E. coli shuttle vector pNW33N using the appropriate restriction enzyme sites. The promoter fragments of ctaB (the promoter exists upstream of ctaB control transcriptional activities of both ctaB and ctaCDEF [40]), the qcrABC and qoxABCD operons, and resD (in addition to the constitutive promoter, ResDE is also controlled by an autoregulated weak promoter directly upstream of resD [57, 58]) were amplified by PCR and cloned into the plasmid pNW33N-gfp. Next, the generated pNW33N-PctaB-gfp, pNW33N-Pqcr-gfp, and pNW33N-Pqox-gfp plasmids were transformed into E. coli TOP10 cells to determine the transcriptional activity of the promoters. The plasmid pNW33N-PresD-gfp was transformed into strain SQR9 to observe the colony fluorescence intensity. The green fluorescent protein (GFP) expression levels of the promoter-gfp fusions in E. coli were monitored by a fluorescence assay (59, 60). All the primers used in this study are shown in Table 2.
Electrophoretic mobility shift assays.The DNA probes (promoter sequences of the qoxABCD, ctaBCDEF, and qcrABC operons) were amplified with fluorescent 6-carboxyfluorescein (FAM)-labeled primers (Table 2). The promoter probes were purified with a Wizard SV gel and PCR clean-up system (Promega, USA) and quantified with a NanoDrop 2000C (Thermo Fisher Scientific, USA). Electrophoretic mobility shift assays (EMSAs) were performed in 20-μl reaction volumes that contained 40 ng probe and 0, 0.28, 0.7, and 1.4 μM ResD protein in a reaction buffer of 50 mM Tris-HCl (pH 8.0), 100 mM KCl, 2.5 mM MgCl2, 0.2 mM dithiothreitol (DTT), 2 μg salmon sperm DNA, and 10% glycerol. After incubating for 30 min at 30°C, the reaction mixture was loaded in a 10% PAGE gel buffered with 0.5× Tris-borate-EDTA (TBE). Gels were scanned with an ImageQuant LAS 4000 mini (GE Healthcare).
DNase I footprint assay.DNase I footprint procedures were modified from previously published methods (61, 62). A 300-bp fragment, encompassing bases −152 to +147 of the qoxABCD operon promoter region, and a 367-bp fragment, encompassing bases −256 to +110 of the ctaBCDEF operon promoter region, were amplified by PCR with the primers qox-PF-6FAM/qox-PR-6FAM and cta-PF-6FAM/cta-PR-6FAM, respectively (Table 2). For each assay, 400 ng of the 6-FAM-labled qoxABCD and ctaCDEF operon promoters was incubated with different amounts of ResD protein in a total volume of 40 μl. After incubating for 20 min at 30°C, 10 μl of a solution containing approximately 0.015 U DNase I (Promega) and 100 nmol of freshly prepared CaCl2 was added, and the reaction mixture was further incubated for 1 min at 30°C. The reaction was stopped by the addition of 140 μl of DNase I stop solution (200 mM unbuffered sodium acetate, 30 mM EDTA, and 0.15% SDS). Samples were first extracted with phenol-chloroform, and then were precipitated with ethanol and the pellets were dissolved in 30 μl of distilled water. The preparation of the DNA ladder, electrophoresis, and data analyses were performed as previously described (63).
Bacterial two-hybrid assay.To perform the bacterial two-hybrid (BTH) assay, we first cloned the coding sequences of the sensor kinase kinB and the qoxABCD operon into the BTH expression vectors. The qoxABCD operon was cloned into the plasmid pKNT25, while the sensor kinase kinB was cloned into the plasmid pUT18 (EuroMedex) (64, 65). Pairwise combinations of plasmids expressing kinB and the qoxABCD operon were cotransformed into the E. coli strain DHM1, which carries a lacZ gene under the control of a cAMP-inducible promoter. The interaction of KinB and QoxABCD would result in the complementation of T25 and T18, which are the catalytic domains of the adenylate cyclase that leads to the production of cAMP, thereby inducing the expression of the lacZ reporter. The transformants were incubated on LB-X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) solid medium supplemented with kanamycin (50 μg/ml) and ampicillin (100 μg/ml) at 30°C for 24 to 48 h. The cells carrying the interacting plasmids became blue in the presence of X-Gal. The empty vectors pKNT25 and pUT18C served as negative controls, whereas pKT25-zip and pUT18C-zip served as positive controls. To quantify the intensity of the interaction, the β-galactosidase activity was determined. The transformants were incubated in LB medium supplemented with ampicillin (100 μg/ml) and kanamycin (50 μg/ml) at 30°C for 48 h in shaker. The cells were permeated using chloroform and 0.01% SDS. β-Galactosidase activity was measured according to Miller (66), and the results are represented as Miller units. All plasmids and primers used in this study are listed in Tables 1 and 2, respectively.
ACKNOWLEDGMENTS
We thank Luo Jun (Nanjing University) for help with ITC experiments.
This work was supported by the National Natural Science Foundation of China (31572214, 31330069 and 31601826), the National Key Basic Research Program of China (973 program, 2015CB150505), and the National Key Research and Development Program (2016YFE0101100 and 2016YFD0200300). R.Z. and Q.S. were also supported by the Key Projects of International Cooperation in Science and Technology Innovation (2016YFE0101100), the 111 Project (B12009), and the Priority Academic Program Development (PAPD) of Jiangsu Higher Education Institutions.
FOOTNOTES
- Received 11 December 2017.
- Accepted 27 January 2018.
- Accepted manuscript posted online 2 February 2018.
Supplemental material for this article may be found at https://doi.org/10.1128/AEM.02744-17.
- Copyright © 2018 American Society for Microbiology.
