DegU and Spo0A Jointly Control Transcription of Two Loci Required for Complex Colony Development by Bacillus subtilis

Biofilm formation is an example of a multicellular process whichdepends on cooperative behavior and differentiation within abacterial population. Our findings indicate that there is acomplex feedback loop that maintains the stoichiometry of theextracellular matrix and other proteins required for complexcolony development by Bacillus subtilis. Analysis of the transcriptionalregulation of two DegU-activated genes that are required forcomplex colony development by B. subtilis revealed additionalinvolvement of global regulators that are central to controllingbiofilm formation. Activation of transcription from both theyvcA and yuaB promoters requires DegU $~$ phosphate, but transcriptionis inhibited by direct AbrB binding to the promoter regions.Inhibition of transcription by AbrB is relieved when Spo0A $~$ phosphateis generated due to its known role in inhibiting abrB expression.Deletion of SinR, a key coordinator of motility and biofilmformation, enhanced transcription from both loci; however, noevidence of a direct interaction with SinR for either the yvcAor yuaB promoter regions was observed. The enhanced transcriptionin the sinR mutant background was subsequently demonstratedto be dependent on biosynthesis of the polysaccharide componentthat forms the major constituent of the B. subtilis biofilmmatrix. Together, these findings indicate that a genetic networkdependent on activation of both DegU and Spo0A controls complexcolony development by B. subtilis.

INTRODUCTION

Top

INTRODUCTION

Bacteria are able to differentiate and coordinate activity intheir offspring to accomplish complex multicellular processes,such as sporulation, biofilm formation, and swarming motility(1, 34, 44). Such multicellular behaviors are optimally exhibitedby undomesticated species of bacteria, and consistent with this,wild isolates of the gram-positive soil bacterium Bacillus subtilisgrow in sessile communities called biofilms that appear to becomplex colonies or air-liquid interface pellicles, whose phenotypiccharacteristics are a result of cooperative behavior and differentiation(1, 43). Unraveling how B. subtilis integrates environmentaland regulatory signals to coordinate the complex decision-makingprocesses that precede and control its diverse multicellularbehaviors has progressed substantially since the first reportsof the biofilm-forming ability of this organism in 2001 (5,17). To date, several transcriptional regulators have been demonstratedto be regulators involved in the formation of a sessile community,and they were recently examined systematically (reference 22and references therein). The contributions of four pairs ofglobal regulators, Spo0A/AbrB, DegS/DegU, SinI/SinR, and SlrR/SlrA,now form the working model for how biofilm formation by B. subtilisis controlled (9, 24).

Biofilm formation, the transition from a free-swimming stateto an adhered state, coincides with the production of an exopolymericmatrix that surrounds the sessile cells (6, 43). In B. subtilistwo operons are essential for this process, an operon containingepsA to epsO (referred to below as the eps operon) and yqxM-sipW-tasA(referred to below as the yqxM operon) (4, 8, 18). The productsof the eps operon direct the synthesis of the polysaccharideconstituent of the extracellular matrix and encode a proteinthat disables the flagellum motor, rendering the cells nonmotile(3). TasA is a protein component of the extracellular matrix,and its correct localization depends on both SipW and YqxM activity(4, 36). SinR inhibits transcription of both the eps and yqxMoperons by direct promoter binding (8, 20), an interaction thatis disrupted by SinI, the antagonist of SinR (2, 20). The pleiotropicregulator AbrB is another negative regulator of biofilm formation(9, 18). AbrB represses transcription of the eps and yqxM operonsvia indirect and direct mechanisms, respectively (9, 18, 37).It additionally directly inhibits the transcription of the yxaABoperon, in which yxaB codes for a putative exopolysaccharidesynthase that is able to restore biofilm formation by a sigBmutant when it is overexpressed (31). Both of these pathwaysare activated by Spo0A through its role in activating transcriptionof sinI and inhibiting transcription of abrB (9).

DegU is a response regulator which needs to be phosphorylatedby its cognate sensor kinase, DegS, to activate biofilm formation(23, 42). Determination that DegU was required for productionof the exopolymer poly- ${gamma}$ -glutamic acid, an extracellular polymerthat can contribute to biofilm formation, indicated that DegUhas further roles during biofilm formation (35). Recently, weand other workers have presented evidence demonstrating thatDegU controls and discriminates between multicellular phenomena,including swarming motility, biofilm formation, and proteaseproduction, via a gradient in its phosphorylation level. Twonovel DegU-regulated genes were identified; in strain NCIB3610yvcA is required for complex colony architecture, and in strainATCC 6051 yuaB is required for pellicle formation (23, 42).These genes encode a putative membrane-bound lipoprotein (YvcA)and a small secreted protein (YuaB), but the exact contributionthat these two proteins make to biofilm formation is not understoodyet.

Here we describe the complex, but strikingly similar, regulatorycontrol of yvcA transcription and yuaB transcription. For bothgenes we show that Spo0A $~$ P exerts positive control through repressingabrB transcription, thus enabling DegU $~$ P-dependent activationof transcription. Direct binding of AbrB to both the yvcA andyuaB promoter regions was confirmed in vitro. We ascribe anindirect regulatory role to SinR since we observed that transcriptionof yvcA and yuaB is elevated in the absence sinR, but we haveshown that SinR does not bind to the yvcA and yuaB promoterregions. The increased level of yvcA and yuaB transcriptionin the sinR mutant was shown to be dependent on production ofthe exopolysaccharide synthesized by the products of the epsoperon, and it is proposed that an intermediate in the intracellularsugar nucleotide cascade that is involved in exopolysaccharidesynthesis has a positive effect on transcription of yvcA andyuaB via a mechanism that so far has not been identified.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

General strain construction and growth conditions. The B. subtilis strains used and constructed in this study aredescribed in Table 1. Escherichia coli strain MC1061 [F' lacI^qlacZM15 Tn10 (tet)] was used for construction and maintenanceof plasmids (Table 2). B. subtilis JH642 derivatives were generatedby transformation of competent cells with plasmids using standardprotocols (19). SPP1 phage transduction for introduction ofDNA into B. subtilis NCIB3610 was conducted as described previously(21). Both E. coli and B. subtilis strains were routinely grownin Luria-Bertani (LB) medium (10 g NaCl per liter, 5 g yeastextract per liter, 10 g tryptone per liter) or MSgg medium (5mM potassium phosphate and 100 mM morpholinepropanesulfonicacid [pH 7.0] supplemented with 2 mM MgCl₂, 700 µM CaCl₂,50 µM MnCl₂, 50 µM FeCl₃, 1 µM ZnCl₂, 2 µMthiamine, 0.5% glycerol, 0.5% glutamate, and appropriate aminoacids at a final concentration of 50 µg ml^–1 [5,8]) at 37°C. When appropriate, antibiotics were used atthe following concentrations: ampicillin, 100 µg ml^–1;chloramphenicol, 5 µg ml^–1; erythromycin, 5 µgml^–1; kanamycin, 25 µg ml^–1; and spectinomycin,100 µg ml^–1. Growth of complex colonies and observationof colony architecture were performed essentially as describedpreviously (5, 42).

View this table:
[in this window]
[in a new window]

TABLE 1. B. subtilis strains used in this study

View this table:
[in this window]
[in a new window]

TABLE 2. Plasmids used in this study

Plasmid construction. The 300-bp region upstream of the yuaB translational start siteand the 240-bp region upstream of the yvcA translational startsite were amplified from NCIB3610 genomic DNA using primer pairsNSW614/NSW615 and NSW408/NSW19, respectively (Table 3). Theresulting PCR fragments were ligated into the BamHI and EcoRIsites located upstream of the lacZ gene in the thrC integrationvector pDG1663 using the restriction sites incorporated in theprimers to generate pNW500 (yuaB) and pNW209 (yvcA) (14) (Table2). Plasmids pNW209 and pNW500 were introduced by selectionfor double-crossover recombination at the thrC locus into JH642and transferred into the chromosome of NCIB3610 by phage transduction(Table 1).

View this table:
[in this window]
[in a new window]

TABLE 3. Primers used in this study

Introduction of sinR into pET28a(+) to generate an N-terminalHis₆ tag fusion was achieved by amplification of the sinR codingregion from the chromosome of NCIB3610 using primers NSW368and NSW369 (Table 3). The resulting PCR product was ligatedinto the NheI and XhoI sites of pET28a(+) to generate pNW90(Table 2). The plasmid was sequenced to ensure that PCR-generatedmutations were not present.

Plasmid pNW518 (Table 2) used to complement the yuaB mutationwas generated by amplifying the yuaB coding sequence, includingthe ribosome binding site, from NCIB3610 genomic DNA using primersNSW626 and NSW645 (Table 3). The resulting PCR fragment wasintroduced into pDR111 using the HindIII and SphI restrictionsites incorporated into the primers. The plasmid was sequencedto ensure that the sequence was correct and was subsequentlyintroduced into JH642 at the amyE locus to generate strain NRS2283(Table 1).

Strain construction. In order to delete sinR, long-flanking homology PCR based ona previously described method (27) was used. The region overlappingthe translational start codon of sinR was amplified using primersNSW148 (5'-CAGGCGCTGAAAACCTTGTATCAACC-3') and NSW149 (5'-ATCACCTCAAATGGTTCGCTGCCAATCAATGTCATCACCTTCC-3'),and the region overlapping the termination stop codon of sinRwas amplified using primers NSW150 (5'-CGAGCGCCTACGAGGAATTTGTATCGGGAGTAGTGCCTGAGCAGAG-3')and NSW151 (5'-GATGCAGCGGCTGCTGAAAAAATC-3'). The translationalstart is indicated by single underlining, and the translationaltermination codon is indicated by double underlining; bold typeindicates sequences homologous to the primer sequences usedto amplify the kanamycin resistance gene. The kanamycin resistancegene was amplified from plasmid pSac-Kan using primers NSW152(5'-CAGCGAACCATTTGAGGTGATAGG-3') and NSW153 (5'-CGATACAAATTCCTCGTAGGCGCTCGG-3').The purified fragments were combined using a sinR 5' PCR product/kanamycinresistance gene/sinR 3' PCR product ratio of 1:2:1 in a PCRmixture containing primers NSW148 and NSW151 and LA Taq (TaKaRa).The resulting PCR fragment was transformed into competent JH642cells, and the resulting colonies were screened by PCR to ensurethat there was double recombination at the sinR locus (NRS1858).The mutated sinR gene was then transferred into NCIB3610 byphage transduction, with selection for kanamycin resistance(NRS1859).

β-Galactosidase assay. Biofilm growth medium (LB medium supplemented with 124 mM potassiumphosphate [pH 7.0], 15 mM ammonium sulfate, 3.4 mM sodium citrate,1 mM MgSO₄, and 0.1% glucose) was inoculated to obtain an opticaldensity at 600 nm (OD₆₀₀) of 0.01 using a preculture grown toan OD₆₀₀ of $~$ 0.5 in LB medium. Cultures were grown in 250-mlErlenmeyer flasks or in glass tubes (1.5 by 15 cm). When required,isopropyl-β-D-1-thiogalactopyranoside (IPTG) was addedafter 60 min of growth. At the time points indicated, the OD₆₀₀was determined and 500 µl of cell suspension was harvestedby centrifugation. Each resulting cell pellet was stored at–20°C for subsequent enzymatic analysis, and the β-galactosidaseassay was performed as described previously (29). On solid mediumcolonies were grown on MSgg medium plates containing 1.5% agarsupplemented with threonine (40 µg/ml) for 18 h. The colonieswere collected from each plate, suspended in 300 µl ofBug Buster Mastermix (Novagen), and incubated for 15 min atroom temperature with gentle shaking. Subsequently, the cellswere vortexed for 10 min after addition of glass beads to furtherdisrupt the cells. The suspension was centrifuged at 16,000x g for 20 min at 4°C to pellet the cell debris. The proteinconcentration of the cell lysate supernatant was determinedusing a Bradford protein assay (Pierce). β-Galactosidaseactivity (in Miller units) was calculated as described aboverelative to the protein concentration in the sample. The valuespresented below are the average β-galactosidase activitiesdetermined for six colonies isolated in three independent experimentsand normalized to the protein concentration (in mg ml^–1);the standard errors of the means are also indicated.

Purification of AbrB and SinR. AbrB and His₆-SinR were purified basically by using previouslydescribed procedures (20, 38). The His₆ tag of His₆-SinR wasremoved using Novagen's thrombin cleavage capture kit.

Electrophoretic mobility shift assay. Protein-DNA gel shift assays were performed with purified SinRand AbrB. The promoter regions assessed were the region frombp –300 to 8 for yuaB (generated using primers NSW617and NSW618), the regions from bp –304 to 50 (generatedusing primers NSW416 and NSW19) and from bp –121 to 50(generated using primers NSW410 and NSW19) for yvcA (P₃₀₀ andP₁₂₀), the region from bp –302 to –1 (generatedusing primers NSW50 and NSW51) for yqxM as a positive control,and (for AbrB) the region from bp –388 to 47 bp (generatedusing primers NSW421 and NSW422) for recA as a negative control.The positions are relative to the translation start codon, andthe A of the ATG codon represented bp 1. ³²P-labeled PCR fragmentswere generated by labeling one primer per primer pair (Table3) prior to PCR amplification using [ ${gamma}$ -³²P]ATP (Perkin Elmer)and T4 polynucleotide kinase (New England Biolabs). Bindingreactions were conducted in 30-µl (total volume) mixturesfor 20 min at room temperature. For SinR the reaction mixturescontained 10 mM Tris-HCl (pH 8.0), 1 mM MgCl₂, 50 mM NaCl, 5%glycerol, 1 mM EDTA, 1 mM dithiothreitol, 50 ng µl^–1bovine serum albumin, 50 ng µl^–1 poly(dI-dC), 0.5to 1 ng ³²P-labeled DNA, and the amounts of SinR indicated below.For AbrB the reaction mixtures contained 20 mM Tris-HCl (pH8.0), 5 mM MgCl₂, 100 mM KCl, 5% glycerol, 0.05% Nonidet P-40,0.5 mM dithiothreitol, 50 ng µl^–1 bovine serum albumin,50 ng µl^–1 poly(dI-dC), 0.5 to 1 ng ³²P-labeledDNA, and the amounts of AbrB indicated below. Reaction samples(15 µl) were loaded onto a 5% polyacrylamide 0.5x Tris-borate-EDTAgel. Gels were run at 100 V, and after drying they were autoradiographedby exposure to a Fujifilm phosphorimaging plate or X-ray film.

RESULTS AND DISCUSSION

Top

RESULTS AND DISCUSSION

YuaB is required for complex colony architecture. The DegU-regulated gene yuaB is required for pellicle formationby strain ATCC 6051 and is postulated to encode a small secretedprotein (23). Transcription of yuaB was shown by Northern blotanalysis to be enhanced in the presence of DegQ, a finding thatindicates that yuaB is transcribed by DegU $~$ P (23). However, deletionof degS from ATCC 6051 and NCIB3610 results in different swarmingmotility phenotypes, indicating that there are differences inthe DegU regulons of these two strains (23, 42). Additionally,ATCC 6051 has recently been found to differ from strain NCIB3610(used in this study) at at least two locations in its genome,at least one of which influences biofilm formation (24, 25).Therefore, it was necessary to establish whether YuaB activatescomplex colony development by strain NCIB3610 and to assessif, like yvcA transcription, yuaB transcription was activatedby DegU $~$ P, as DegU is known to have regulatory activity in bothits nonphosphorylated and phosphorylated states (10). Highlyconsistent with the role of DegU and YuaB during pellicle formationin strain ATCC 6051, an NCIB3610-derived strain with a deletionin the single-gene operon yuaB (NRS2097) showed impaired complexcolony architecture (Fig. 1A). Since the impact on complex colonyarchitecture was greater than the impact on deletion of yvcAobserved (42), this may indicate that yuaB encodes a proteinrequired earlier in the complex colony development process thanYvcA (Fig. 1A). The deletion phenotype was efficiently complementedby ectopic expression of the yuaB coding sequence under thecontrol of a heterologous promoter from the amyE locus (Fig.1B). While partial complementation was observed in the absenceof IPTG, full complementation of the yuaB mutation was seenafter addition of IPTG at concentrations of 10 and 100 µM.How YuaB functions is not known yet, but one hypothesis is that,like the secreted protein TasA (4), YuaB forms an additionalstructural component of the extracellular matrix. Alternatively,YuaB may form part of a feedback signaling cascade to regulategene expression required for complex colony development by B.subtilis.

View larger version (59K):
[in this window]
[in a new window]

FIG. 1. Complex colony architecture requires DegU-dependent yuaB transcription. (A) Complex colony morphology of wild-type strain NCIB3610, yuaB mutant NRS2097, and yvcA mutant NRS1390 after 48 h of incubation at 37°C on MSgg agar. (B) Complementation of yuaB mutant NRS2097 through ectopic expression of yuaB under control of a heterologous promoter (NRS2299). Strains were grown in the presence of IPTG as indicated at the top for 48 h at 37°C on MSgg agar. (C) Transcription of yuaB (PyuaB-lacZ) in wild-type strain NRS2052 (WT), degU mutant NRS2224, and the degU mutant complemented with the P_IPTG-degU-D56N (NRS2056) and P_IPTG-degU (NRS2226) constructs. The IPTG concentrations used were 0, 5, and 25 µM for the P_IPTG-degU construct (+WT0, +WT5, and +WT25, respectively) and 0 and 50 µM for the P_IPTG-degU-D56N construct (+noP0 and +noP50, respectively). The bars indicate the β-galactosidase activities and the error bars indicate the standard errors of the means obtained for stationary-phase cells in three to five independent experiments.

In line with previous findings for yvcA (42), transcriptionof yuaB was activated by DegU $~$ P when lacZ was used as an in vivoreporter of transcription using a 300-bp yuaB promoter region.Deletion of degU resulted in an eightfold decrease in yuaB transcriptionwhich could not be complemented by ectopic expression of a mutantallele of degU that cannot be phosphorylated due to a pointmutation in the amino acid that is phosphorylated by DegS (degU-D56N)(Fig. 1C) (42). In contrast, ectopic expression of the wild-typedegU gene under the control of the IPTG-inducible promoter P_hy-spank(42) resulted in wild-type levels of yuaB expression after additionof 5 µM IPTG and levels slightly higher than the wild-typelevels after addition of 25 µM IPTG (Fig. 1C). These dataindicate that, like the overall complex colony development process,transcription of yuaB is activated by DegU $~$ P (23, 42).

AbrB represses transcription of yvcA and yuaB. Only a limited number of genes that are known to be directlyregulated by DegU have been characterized with respect to additionalregulatory control, but what has become apparent from thesestudies is that multiple global regulators are involved (16,33). For example, transcription of aprE, in addition to beingactivated by DegU, is controlled by SinR, AbrB, and Spo0A (2,12). These regulators are all key players in the control ofcomplex colony development by B. subtilis (5, 9, 17, 18, 22).Therefore, it was hypothesized that SinR, SinI, AbrB, and Spo0Amay have a role in regulating the novel DegU-activated loci,yvcA and yuaB, that are also required for complex colony development(Fig. 1A). This possibility was addressed using mutant strainscarrying lacZ as an in vivo reporter of either yvcA or yuaBtranscription. For yvcA the 240-bp sequence upstream of thetranslational start site was used to drive lacZ transcriptionas this region of DNA has been defined as the optimal minimalyvcA promoter region (P₂₄₀) (see Fig. S1 in the supplementalmaterial).

Introduction of a mutant allele of spo0A encoding a versionof Spo0A that cannot be phosphorylated resulted in fivefold-decreasedyvcA transcription during stationary phase (P < 0.01) (Fig.2B). This finding is consistent with the role of Spo0A $~$ P in activatingbiofilm formation (5, 17). It was subsequently determined thatthe lack of yvcA transcription observed in the spo0A mutantwas mediated through its known role in inhibiting abrB expression(17, 39). Consistent with this, yvcA transcription was observedto be constitutive in an abrB mutant strain, and the inhibitionof transcription observed in the spo0A mutant was bypassed inthe abrB spo0A double mutant strain (Fig. 2A and 2B). The samephenomenon was observed when transcription from the yuaB promoterwas examined. Deletion of spo0A resulted in an 11-fold (P <0.01) decrease in transcription. In the abrB and spoOA mutants,the level of transcription was four- to fivefold higher (P <0.01) than the wild-type level, respectively (Fig. 2C). Thesedata indicate that both yvcA and yuaB are regulated indirectlyby Spo0A $~$ P through AbrB. Consistent with these findings, neithergene was identified as a direct target of Spo0A using chromatinimmunoprecipitation experiments (30), but both genes were determinedby genome-wide transcriptome analysis to be members of the largerspo0A regulon (11).

View larger version (15K):
[in this window]
[in a new window]

FIG. 2. Transcription of yvcA and yuaB is activated by Spo0A $~$ P through abrB inhibition. (A) Transcription of yvcA (PyvcA-lacZ) in the wild-type (NRS1608) and abrB mutant (NRS1644) backgrounds. The symbols indicate the β-galactosidase activities (in Miller units) (filled symbols) and the growth (open symbols) expressed as OD₆₀₀ for the wild-type strain (squares) and abrB mutant (diamonds) strain at different times. (B) Transcription of yvcA (PyvcA-lacZ) in the wild-type strain (NRS1608) (WT), the spo0A mutant (NRS1631), the abrB mutant (NRS1644), and the spo0A abrB mutant (NRS2067). The bars indicate the β-galactosidase activities (in Miller units) and the error bars indicate the standard errors of the means obtained for exponential-phase cells (120 min) (open bars) and for stationary-phase cells (360 min) (filled bars). (C) Transcription of yuaB (PyuaB-lacZ) in the wild-type strain (NRS2052), the spo0A mutant (NRS2053), the abrB mutant (NRS2228), and the spo0A abrB mutant (NRS2233). The bars indicate the β-galactosidase activities (in Miller units) and the error bars indicate the standard errors of the means obtained for stationary-phase cells (360 min).

AbrB binds to the promoter region of yvcA and yuaB. To establish that the AbrB-mediated repression of transcriptionwas a direct regulatory event, DNA retardation assays with theyvcA and yuaB promoter regions and purified AbrB were performed.Using yqxM as a positive control (37) and the recA promoterregion as a negative control (15), a direct interaction wasdemonstrated (Fig. 3). Addition of AbrB at a concentration of150 nM was sufficient to initiate a promoter DNA-AbrB interaction,while complete yvcA and yuaB promoter binding was observed inthe presence of 250 nM AbrB (Fig. 3). AbrB binding was not detectedwhen only the 120-bp sequence upstream of the translationalstart site for yvcA, a region that also did not show any DegU $~$ P-activatedtranscription when it was fused with lacZ (see Fig. S1 in thesupplemental material), was used. These data indicate that AbrBbinds to the region between bp –120 and –240 upstreamof the yvcA translational start site and binds within the 300-bpregion upstream of the translation start site of yuaB, and theyadd yvcA and yuaB to the growing list of AbrB-inhibited targetsthat need to be derepressed for complex colony development tooccur. These data highlight the key role that AbrB plays incontrolling this multicellular behavior exhibited by B. subtilis(9, 17, 18, 31).

View larger version (37K):
[in this window]
[in a new window]

FIG. 3. AbrB is a direct repressor of yvcA and yuaB transcription. An electrophoretic mobility shift assay was performed using ³²P-labeled yvcA₃₀₀, yvcA₁₂₀, yuaB, yqxM, and recA promoter fragments incubated with AbrB at the concentrations indicated. +ve, positive; –ve, negative.

Deletion of sinR enhances transcription of yvcA and yuaB. Consistent with the role of SinR as a repressor of complex colonyarchitecture (20), deletion of sinR resulted in a fourfold increasein transcription from the yvcA promoter during exponential growth(Fig. 4A and 4B) and a fourfold (P < 0.01) increase in transcriptionfrom the yuaB promoter throughout growth compared with the wild-typestrain (Fig. 5A). SinI is the antagonist of SinR (2), and consistentwith this, deletion of sinI resulted in a twofold decrease intranscription for both the yvcA and yuaB promoters comparedwith the wild type (Fig. 4B and 5A). As expected, deletion ofboth sinR and its antagonist sinI in combination resulted inan increase in transcription similar to that observed for thesingle sinR deletion (Fig. 4B and 5A). We wanted to ensure thatthe fourfold-enhanced transcription was not the result of inaccurateassessment of the cell number (and therefore gene transcription)in the sinR and sinIR mutant strains as shaking cultures ofthe sinR and sinIR mutant strains do not have a uniform opticaldensity due to overexpression of components of the biofilm matrix(20). We did not think that this was likely as we used an LBmedium-based growth medium for our experiments, in which thefinal transcriptional level of the eps operon in the absenceof the repressor sinR was suppressed (data not shown). Thus,in this growth medium, the cells still clumped but at a lowerlevel than in MSgg medium (data not shown), and therefore, thegrowth could be followed (Fig. 2A and 4A). However, in orderto eliminate any doubt, the level of expression from the yvcApromoter in the presence and absence of sinR was calculatedafter preparation of soluble-protein cell lysates. This allowednormalization of the β-galactosidase activity to proteinlevels. Using this approach, a 4.5-fold-higher level of yvcAtranscription in the sinR mutant was still observed, indicatingthat the recorded values were not the result of inaccurate assessmentof the cell number (see Fig. S2 in the supplemental material).Our findings indicate that SinR inhibits transcription of boththe yvcA and yuaB DegU-activated promoters. To establish thatit was a direct interaction, we used electrophoretic mobilityshift assays with purified SinR (Fig. 5C). Using radiolabeledyvcA and yuaB promoter regions, we were unable to demonstrateany interaction of SinR with either promoter region in the nanomolarrange needed to bind to the yqxM promoter, a known SinR target(Fig. 5C) (8, 20). Therefore, in contrast to the known roleof SinR in repressing eps and yqxM operon transcription (4,8, 20), these findings do not support a model in which the inhibitionof yvcA and yuaB transcription by SinR is directly mediated.

View larger version (15K):
[in this window]
[in a new window]

FIG. 4. Deletion of sinR enhances transcription of yvcA. (A) Transcription of yvcA (PyvcA-lacZ) in the wild-type background (NRS1608) and the sinR mutant (NRS1628). The symbols indicate the β-galactosidase activities (in Miller units) (filled symbols) and the growth (open symbols) expressed as OD₆₀₀ for the wild-type strain (squares) and sinR mutant strain (diamonds) at different times. (B) Transcription of yvcA (PyvcA-lacZ) in the wild-type background (NRS1608), the sinR mutant (NRS1628), the sinI mutant (NRS2083), and the sinIR mutant (NRS2084). The bars indicate the β-galactosidase activities (in Miller units) and the error bars indicate the standard errors of the means obtained for exponential-phase cells (120 min) (open bars) and stationary-phase cells (360 min) (filled bars) for three to five experiments.

View larger version (27K):
[in this window]
[in a new window]

FIG. 5. SinR is not a direct repressor of yuaB transcription. (A) Transcription of yuaB (PyuaB-lacZ) in the wild-type background (NRS2052) (WT), the sinR mutant (NRS2223), the sinI mutant (NRS2055), the sinIR mutant (NRS2230), the epsG mutant (NRS2054), the epsG sinR mutant (NRS2231), and the P_IPTG-eps operon strain (NRS2232) with no IPTG (+Eps0) and 25 µM IPTG (+Eps25). The bars indicate the β-galactosidase activities (in Miller units) and the error bars indicate the standard errors of the means obtained for stationary-phase cells in three to five independent experiments. (B) Complex colony architecture can be restored to the epsA::pBL584( ${Phi}$ [Pspachy-epsA-O]) strain (P_IPTG-epsA-O) (NRS1685) in the presence of 30 µM IPTG after incubation for 24 h at 37°C on MSgg medium. (C) Electrophoretic mobility shift assay using ³²P-labeled yvcA₃₀₀, yqxM, and yuaB promoter fragments incubated with SinR at the concentrations indicated. (D) Transcription of yuaB (PyuaB-lacZ) in the wild-type (NRS2052), degU mutant (NRS2224), sinR mutant (NRS2223), epsG mutant (NRS2054), and epsG sinR mutant (NRS2231) backgrounds. The bars indicate the β-galactosidase activities (in Miller units) and the error bars indicate the standard errors of the means normalized to the total protein concentration (in mg ml^–1) determined using colonies grown on MSgg medium for 18 h in three independent experiments.

Synthesis of the exopolysaccharide is required for enhanced transcription. In the absence of a direct interaction of SinR with the yvcAand yuaB promoters, we postulated that the increase in transcriptionof yvcA and yuaB may be stimulated by the extracellular matrixor an intermediate in the biosynthetic process that is overproducedin both the sinR and sinIR mutants. To test this hypothesis,the level of transcription from the yvcA and yuaB promoterswas measured in the presence of a mutation in the eps operonwhich prevents exopolysaccharide biosynthesis (20). In an epsGsinR mutant strain the yvcA transcription was equivalent tothat of the wild-type strain (data not shown), and the yuaBtranscription was 1.7-fold lower than that of the wild-typestrain (Fig. 5A). These findings are consistent with the hypothesisthat exopolysaccharide production stimulates transcription ofthese two DegU-activated loci. These experiments were conductedusing cells grown in liquid LB medium-based growth medium, andwe wanted to ensure that the same pattern of gene expressionoccurred under complex colony development conditions; therefore,we harvested colonies from solid MSgg medium after 18 h of growth.The level of β-galactosidase activity was normalized tothe total amount of protein extracted from the colony. We observeda 2.5-fold increase (P < 0.0001) in yuaB transcription inthe absence of sinR compared with the transcription in the wild-typestrain (Fig. 5D) and confirmed that the increase in transcriptionwas dependent on the production of the exopolymeric matrix sincedisruption of epsG in the sinR background resulted in loss ofthe enhanced level of transcription (Fig. 5D).

To further investigate the hypothesis that synthesis of thepolysaccharide component of the biofilm matrix acted as a signalresponsible for enhancing yvcA and yuaB transcription, transcriptionof the eps operon was uncoupled from native regulation by replacingthe eps operon promoter with a construct that enabled transcriptionof the eps operon from an IPTG-inducible promoter. Substitutionof the promoter region allowed wild-type colony morphology tobe exhibited in the presence of 30 µM IPTG (Fig. 5B).Using the P_IPTG-eps operon strain, the level of yuaB transcriptionwas measured in the presence and absence of IPTG. Without additionof IPTG to cells carrying the P_IPTG-eps operon construct, thetranscription of yuaB during early stationary phase was notsignificantly different from that measured for the wild-typestrain, but addition of 10 µM IPTG resulted in a smallbut statistically significant increase in transcription (1.7-fold;P < 0.01 [data not shown]). Addition of 25 µM IPTGincreased transcription of yuaB 3.5-fold (P < 0.05) duringstationary phase to levels comparable to those observed forthe sinR mutant strain (Fig. 5A). These findings are consistentwith the DNA retardation data which indicated that SinR doesnot bind to the yvcA and yuaB promoters and could explain whyyvcA and yuaB were not identified as SinR-regulated genes ingenome-wide transcriptional profiling studies since those analyseswere performed using a sinR mutant strain carrying a mutationin the eps operon (8). In conclusion, our data indicate thateither a product of the eps operon, the biofilm matrix itself,or some unidentified regulatory step that requires matrix productionenhances transcription of yvcA and yuaB in the sinR mutant strain.

The exopolysaccharide is not an extracellular signal triggering enhanced yuaB transcription. It is possible that the exopolysaccharide synthesized by theproducts of the eps operon could enhance transcription of yvcAand yuaB through recognition by a receptor on the cell surface,which in turn could trigger a signaling cascade within the cell.If this were the case, then overproduction of the exopolysaccharidewould be detected for a strain that does not overproduce theexopolysaccharide when it is assessed in coculture experiments.To test this possibility, the level of transcription from theyuaB promoter fusion was measured for the wild-type strain thatwas cocultured at a 1:1 ratio with either the wild-type strainor the sinIR mutant strain that lacked the yuaB reporter fusion.It was established that the increased level of exopolysaccharideproduced by the sinIR mutant did not result in enhanced transcriptionof yuaB in the wild-type strain, indicating that the exopolysaccharidedid not function as an extracellular signaling molecule (datanot shown). It is not known yet how overproduction of the extracellularpolysaccharide enhances transcription of the yvcA and yuaB loci;however, one hypothesis favored by us is that carbohydrate intermediatesin the polysaccharide biosynthetic process are detected intracellularlyand used to mediate signal transduction processes. Detectionof metabolic intermediates as a means to control gene expressioncan occur through the activity of trans-acting ribozymes, suchas the glucosamine-6-phosphate-activated ribozyme encoded byglmS (26, 45).

Conclusion. AbrB and SinR are known for their role in inhibiting transcriptionof the eps and yqxM operons that are required for biosynthesisof the extracellular matrix of the B. subtilis biofilm (9, 20).Here we found that transcription of the DegU-activated genesyvcA and yuaB is also regulated by AbrB and SinR. While we wereable to clearly demonstrate that the inhibition exerted by AbrBwas directly mediated (Fig. 3), we surprisingly also demonstratedthat the inhibition exerted by SinR was indirectly mediatedthrough the exopolysaccharide synthesized by the protein productsof the eps operon (Fig. 5). These findings indicate that a complexfeedback loop is established during complex colony developmentby B. subtilis to ensure that the stoichiometry of yvcA andyuaB transcription is kept in balance with that of the eps andyqxM operons (Fig. 6). The dual requirements of DegU and Spo0Aactivation prior to complex colony development are comparableto the regulatory network required to activate exoprotease productionby B. subtilis. Exprotease production occurs in a small subpopulationof cells that reach threshold levels of DegU $~$ P and Spo0A $~$ P (40,41). Therefore, our findings add complex colony developmentto the growing list of multicellular behavioral phenomena exhibitedby B. subtilis that are subjected to an "AND-gating geneticcircuit" that depends on the activation of both DegU and Spo0A(40, 41). Future work will investigate how this regulatory loopis maintained and whether fluctuations in the absolute levelsof DegU and Spo0A activation are used as a mechanism to coordinateentry into the different multicellular behaviors that are exhibitedby B. subtilis.

View larger version (26K):
[in this window]
[in a new window]

FIG. 6. A complex regulatory network controls biofilm formation. The diagram is a simplified diagram depicting the network that controls yvcA and yuaB transcription during complex colony development (5, 8, 9, 17, 20). Open arrowheads indicated changes in phosphorylation, filled arrowheads indicate activation of transcription, and T bars indicate inhibition of transcription.

ACKNOWLEDGMENTS

This work was supported by the Biotechnology and BiologicalSciences Research Council (grants BB/C520404/1 and BB/E001572/1).

We thank The Sequencing Service (College of Life Sciences, Universityof Dundee, Dundee, Scotland) (www.dnaseq.co.uk) for DNA sequencing.We are very grateful to M. Strauch (University of Maryland)for providing the AbrB-overproducing E. coli strain and forhis advice concerning AbrB purification. We thank D. Kearns(University of Indiana) for providing the sinI and sinIR strainsand for helpful discussions, K. Kobayashi (NIST, Japan) forproviding the yuaB mutant strain, Taryn Kiley (University ofDundee) for providing the sinR::kan strain, and B. Terra andB. Lazazzera (University of California, Los Angeles) for providingstrain BAL2340 prior to publication and for helpful discussions.

FOOTNOTES

* Corresponding author. Mailing address: Division of Molecular and Environmental Microbiology, College of Life Sciences, MSI/WTB/JBC Complex, University of Dundee, Dundee, United Kingdom DD1 5EH. Phone: 44(0)1382 386335. Fax: 44(0)1382 386375. E-mail: n.r.stanleywall{at}dundee.ac.uk

Published ahead of print on 31 October 2008.

Supplemental material for this article may be found at http://jb.asm.org/.

REFERENCES

Top