Environmental Calcium Initiates a Feed-Forward Signaling Circuit That Regulates Biofilm Formation and Rugosity in Vibrio vulnificus

The second messenger c-di-GMP is a key regulator of bacterial physiology. The V. vulnificus genome encodes nearly 100 proteins predicted to make, break, and bind c-di-GMP. However, relatively little is known regarding the environmental signals that regulate c-di-GMP levels and biofilm formation in V. vulnificus. Here, we identify calcium as a primary environmental signal that specifically increases intracellular c-di-GMP concentrations, which in turn triggers brp-mediated biofilm formation. We show that PAPS, a metabolic intermediate of the sulfate assimilation pathway, acts as a second messenger linking environmental calcium and sulfur source availability to the production of another intracellular second messenger (c-di-GMP) to regulate biofilm and rugose colony formation, developmental pathways that are associated with environmental persistence and efficient bivalve colonization by this potent human pathogen.

B acteria can exist as free-living, planktonic cells or in surface-attached, multicellular biofilm communities that are composed of bacteria encased in a matrix of polysaccharides, DNA, and proteins (1). Biofilms are more resistant than their planktonic counterparts to environmental stressors such as antibiotic exposure, oxidative agents, ionic stress, and desiccation. The process of biofilm formation is tightly controlled at multiple stages, from approaching and probing a surface, to irreversible attachment and maturation, to biofilm dispersal (2). The second messenger c-di-GMP is a key biofilm regulatory molecule (3), and bacteria can sense diverse environmental signals and modulate their intracellular c-di-GMP levels accordingly. It is synthesized by diguanylate cyclases (DGCs) that bear a GGDEF domain and is degraded by phosphodiesterases (PDEs) that harbor EAL or HD-GYP domains (4). The GGDEF and EAL/HD-GYP domains of DGC and PDE are often associated with sensor domains that regulate their function in response to external cues such as oxygen level, light, nitric oxide, temperature, nutrient limitation, and surface contact (5). Additionally, DGC and PDE activity can be regulated at the level of gene expression. The downstream effects of c-di-GMP are mediated by effectors (proteins or nucleic acids such as riboswitches) that bind the molecule and elicit a change in gene expression or protein function (6).
In Vibrio vulnificus, elevated levels of intracellular c-di-GMP trigger biofilm formation by inducing expression of the brp-encoded exopolysaccharide (EPS) (7,8). Excessive production of EPS and matrix proteins promotes development of the rugose phenotype, which is characterized by a distinct wrinkly colony morphology, and increased biofilm formation and stress resistance (9)(10)(11). brp expression is also dependent on the regulators BrpR and BrpT (7,12). BrpR shares homology with VpsR of Vibrio cholerae, which has been shown to bind c-di-GMP (13,14). BrpT binds directly to promoter regions upstream of brpA and brpH, but unlike its V. cholerae homologue VpsT, it does not depend on first binding c-di-GMP to do so (12,15). However, the expression of brpT is regulated by BrpR (12). It was also recently shown that c-di-GMP, via BrpT, regulates the expression of the cabABC operon (16), which encodes a system for the secretion of a calcium binding matrix protein CabA that is required for biofilm and rugose colony formation (17).
The V. vulnificus genome encodes nearly 100 proteins predicted to synthesize, degrade, and bind c-di-GMP (18,19), but relatively little is known regarding the environmental signals that regulate c-di-GMP levels and biofilm formation in response to changing environmental conditions. Vibrio vulnificus and the bivalves (oysters) it colonizes are autochthonous to estuary ecosystems (20,21). These partially enclosed bodies of water are in constant flux due to the varying flows of freshwater (rainfall and snowmelt) and seawater (changing tides) that enter, mix with, and exit the water column (22). The salinity can range from 5 to 30 ppt, vary between estuaries, and change daily depending on the tides, weather, and anthropogenic factors. It was previously reported that calcium increased phase variation and biofilm formation in V. vulnificus (23,24), but the mechanism remained nebulous. Here, we identify calcium as a primary environmental signal that specifically increases intracellular c-di-GMP concentrations, which in turn triggers subsequent brp-mediated biofilm formation. A screen for transposon (Tn) mutants that failed to respond to increased calcium revealed that sulfate assimilation is genetically linked to brp expression. We show that a specific intermediate of the sulfate assimilation pathway, 3=phosphoadenosine 5=-phosphosulfate (PAPS), is required for calcium-induced brp expression. Moreover, PAPS production was also separately required for development of the physiologically distinct rugose phenotype. These results support the notion that PAPS is an intracellular second messenger linking environmental calcium and sulfur source availability to the production of another intracellular second messenger (c-di-GMP) to regulate the physiological responses of biofilm formation and rugosity, phenotypes that underpin the evolution of V. vulnificus as a successful environmental organism.

RESULTS
Environmental calcium regulates brp expression by moderating the intracellular c-di-GMP level. Our previous investigations indicated that brpA expression (monitored using a P brpA lacZ reporter construct) was low in wild-type cells grown in LB. Conversely, the same reporter strain exhibited high-level induction when grown on marine agar (MA) (Fig. 1A), a medium that better reflects the composition of seawater. To identify the component(s) of MA that triggered brpA expression, the reporter strain was grown on LB that was systematically supplemented with the major mineral constituents of MA. While the major mono-and divalent cations present in seawater (Fig. 1B) had a negligible effect on brp expression when added to LB at typical seawater concentrations, strikingly, the addition of CaCl 2 to 10 mM strongly induced P brpA expression (Fig. 1C). These results suggested that calcium specifically regulated brp expression. Disruption of either brpF or brpJ (which encode a glycosyltransferase and a transport protein, respectively) did not affect calcium-induced brp expression (see and a breakdown of the concentration (mM) of the major sea salts (right). Data are from the National Oceanographic Data Center (https://www.nodc.noaa.gov/). (C) P brpA gfp expression in V. vulnificus grown in LB or LB supplemented with the major sea salts to concentrations found in seawater. Shown on the right are wild-type V. vulnificus cells bearing a P brpA lacZ reporter growing on LB agar-X-Gal plates containing 50 mM MgCl 2 (LB Mg ) or 10 mM CaCl 2 (LB Ca ). (D) Fluorescent imaging of V. vulnificus biofilms growing in LB or LB containing 5 or 10 mM CaCl 2 (LB Ca-5 and LB Ca-10 ). An axis indicator is shown in the bottom left corner of each panel. Statistically significant differences among the samples (P Ͻ 0.001 as determined by one-way ANOVA followed by pairwise comparisons with a Bonferroni adjustment) are indicated by different symbols above each bar.
Calcium Triggers Biofilm Formation in V. vulnificus ® Fig. S1 in the supplemental material), suggesting little or no feedback inhibition of P brpA expression due to incomplete polysaccharide biosynthesis or failed export of a completed polysaccharide. Notably, while increasing Ca 2ϩ concentrations enhanced V. vulnificus biofilm formation (Fig. 1D), it repressed V. cholerae biofilm formation under the same conditions (see Fig. S2 in the supplemental material). These results indicated that environmental calcium can have different effects on biofilm formation by closely related species.
Expression of the brp locus is regulated by the intracellular concentration of c-di-GMP (7,12). To assess if the increase in brp expression in response to extracellular calcium correlated with an elevation in intracellular c-di-GMP levels, wild-type cells were grown in LB or LB supplemented with 50 mM MgCl 2 or 15 mM CaCl 2 (LB Mg and LB Ca , respectively), and biofilm formation and the intracellular c-di-GMP levels were quantified. Biofilm formation was low for cells grown in LB or LB Mg but increased 7-fold in LB Ca ( Fig. 2A). The c-di-GMP level in cells that were grown in LB Mg was half that of cells grown in LB (Fig. 2B). Conversely, the c-di-GMP level in cells that were grown in LB Ca was 60% higher than that of cells grown in LB. Collectively, these data suggested that the intracellular c-di-GMP level in V. vulnificus increased in response to environmental calcium, which in turn regulated brp expression.
Cysteine biosynthesis and calcium-induced brp expression are genetically linked. A V. vulnificus transposon mutant library was screened for the loss of P brpA lacZ expression under calcium-inducing conditions to determine how an increase in exogenous calcium concentration resulted in increased brp expression (see Table S3 in the supplemental material). Multiple insertions mapped to brpT. This was anticipated since brpT is absolutely required for brp expression (7). Of particular interest was an insertion that mapped to cysD (aot11_12580), the gene coding for sulfate adenylyltransferase, which participates in cysteine biosynthesis (25). Our previous transcriptome sequencing (RNA-seq) analyses (16) indicated that cysD transcript levels increased when intracellular c-di-GMP levels were elevated but failed to do so in a brpT mutant. These data suggested that cysD expression was positively regulated by c-di-GMP and dependent on BrpT.
A cysD null mutant bearing the P brpA lacZ reporter was constructed and grown on LB Ca agar to verify that a functional CysD was indeed required for calcium-induced brpA expression. P brpA expression was elevated in wild-type cells but not in the cysD mutant, and this defect could be complemented by expressing an ectopically integrated copy of cysD (Fig. 3A). Accordingly, P brpA lacZ expression was low in LB but increased 4-fold in wild-type cells grown in LB Ca (Fig. 3B). In contrast, a mutation in cysD prevented calcium induction of brpA expression, and this could be complemented by ectopic expression of cysD. Thus, a functional CysD was necessary for P brpA expression in response to calcium.
The sulfate assimilation metabolic intermediate PAPS mediates calciuminduced brp expression. Disruption of cysD is anticipated to render cells auxotrophic for cysteine. However, the Tn libray screen that unearthed cysD was conducted on LB agar, which contains up to 0.3 mM cysteine (26). Previous reports of mutant screens on LB agar have resulted in the recovery of cysteine-auxotrophic mutants (27). To determine if the V. vulnificus cysD mutant was indeed auxotrophic, the ΔcysD P brpA lacZ reporter strain was grown on minimal medium containing (MM cys ) or lacking cysteine (MM). The wild-type strain grew well on both types of medium (see Fig. S3 in the supplemental material), while the mutant grew well on MM cys but failed to grow on MM. This confirmed that the cysD mutant was indeed auxotrophic for cysteine but that cysteine was not limiting for growth under the conditions used for Tn mutagenesis. Notably, these results suggested that the failure of a cysD mutant to induce brp expression in response to elevated calcium was unlikely to be rooted in its inability to synthesize cysteine itself. Rather, they implicated an intermediate in cysteine biosynthesis as requisite for calcium-induced P brpA expression.
The cysDNC operon participates in assimilatory sulfate activation in bacteria (see Fig. S4 in the supplemental material) (25). Environmental sulfate is transported into the cell, where CysD, together with CysN, then uses ATP and GTP to convert the acquired sulfate to adenosine 5=-phosphosulfate (APS). CysC, an adenylyl-sulfate kinase, uses another molecule of ATP to convert APS to 3=-phosphoadenosine 5=-phosphosulfate (PAPS), better known as activated sulfur. CysH is a sulfotransferase that converts PAPS to sulfite and adenosine 3=,5=-bisphosphate (PAP). Sulfite is converted to sulfide by the action of cysI and cysJ, which can then participate in sulfurylation reactions or be utilized in cysteine biosynthesis. To identify the sulfate assimilation intermediate regulating P brpA expression, null mutations that systematically targeted specific steps of de novo cysteine biosynthesis were constructed. Deletions of cysD, cysN, cysC, and cysH were generated in the P brpA lacZ reporter strain, and the mutants were assayed for growth in medium that included or lacked cysteine. All of the strains grew equally well on MM cys , but the mutants failed to grow on MM (Fig. S3), confirming their auxotrophic phenotype. The same strains were grown in LB Ϯ 10 mM CaCl 2 (ϩCa) for quantification of P brpA lacZ expression. Statistically significant differences among the samples (P Ͻ 0.001 as determined by one-way ANOVA followed by pairwise comparisons with a Bonferroni adjustment) are indicated by different symbols above each bar.
Calcium Triggers Biofilm Formation in V. vulnificus ® Calcium-induced P brpA lacZ expression was monitored in each mutant background. P brpA expression in the ΔcysD, ΔcysC, and ΔcsyH ΔcysC mutants was reduced relative to wild type on LB Ca and LB that lacked calcium ( Fig. 4A and B). Conversely, P brpA expression increased in the ΔcysH strain on LB Ca and LB, and cellular c-di-GMP levels of the ΔcysH strain were nearly twice that of the wild type and the ΔcysC, ΔcysD, and ΔcysC ΔcysH mutants on LB and LB Ca . These results suggested that mutations that interfered with the conversion of sulfate to PAPS (ΔcysD and ΔcysC) inhibited calcium-induced P brpA expression, while a mutation that favored the accumulation of PAPS by restricting its conversion to sulfite (ΔcysH) had increased intracellular c-di-GMP levels and increased P brpA expression.
Deletion of cysD does not prevent c-di-GMP-mediated P brpA expression. To determine if brp expression was absolutely dependent on PAPS production, the diguanylate cyclase, DcpA, was expressed in wild-type and ΔcysD cells to increase the intracellular c-di-GMP concentration and P brpA lacZ expression was monitored. An equally robust increase in P brpA expression was observed in both the wild-type and ΔcysD strains (Fig. 5). This suggested that elevated intracellular c-di-GMP levels could bypass the input of calcium, cysD, and PAPS in regulating brp expression.
PAPS production is required for rugosity. The extent of rugose colony development in response to increased intracellular c-di-GMP concentrations in the wild-type , and the same strains were grown in LB Ϯ 10 mM CaCl 2 for quantification of P brpA lacZ expression (middle) and the intracellular c-di-GMP concentration (bottom) in the WT and cys mutants. Statistically significant differences among the samples (P Ͻ 0.001 as determined by one-way ANOVA followed by pairwise comparisons with a Bonferroni adjustment) are indicated by different symbols above each bar. and ΔcysD strains did not correlate with their comparably equal levels of P brpA expression. Rugose colony formation and an associated increase in surface area were evident in wild-type cells expressing DcpA ( Fig. 6A and B). A lesion in brpJ abrogated this phenotype, indicating that rugose colony development was dependent on the production of the brp polysaccharide. While there was little effect on rugose colony  Calcium Triggers Biofilm Formation in V. vulnificus ® development when cysH was deleted, the ΔcysD and ΔcysC strains failed to become rugose when intracellular c-di-GMP levels were increased. Moreover, P brpA expression increased in a ΔcysD background when intracellular c-di-GMP levels were elevated (Fig. 5), yet colonies failed to become rugose, and deletion of cysD abrogated pellicle formation at the air-liquid interface relative to wild-type cells (Fig. 6C and D). These results suggested that the ability to produce PAPS (which is reduced in ΔcysD and ΔcysC relative to wild-type and ΔcysH cells) was required for rugose colony development and likely pellicle formation in a manner that was distinct from its role in regulating P brpA expression.
c-di-GMP and BrpT regulate the expression of genes in the sulfate assimilation pathway. Our previous comparative transcriptomic analyses under unaltered and elevated cellular c-di-GMP conditions (12) suggested two groupings for the genes of sulfate assimilation: the expression of those involved in the conversion of sulfate to PAPS (cysDNC) increased when intracellular c-di-GMP concentrations were elevated, while those required for the production of cysteine from PAPS (cysHIJ) were downregulated (see Table S4 in the supplemental material). Moreover, the expression of cysDNC appeared to be positively regulated by BrpT. To confirm this, quantitative reverse transcription-PCR (qRT-PCR) was performed on RNA isolated from wild-type and ΔbrpT cells expressing DcpA or carrying the empty vector. In agreement with our RNA-seq data, the expression of cysD was strongly induced in wild-type cells under elevated intracellular c-di-GMP conditions (Fig. 7), while the expression of cysJ decreased 2-fold (see Fig. S5 in the supplemental material). Moreover, the c-di-GMPmediated increase in cysD expression was negated in a brpT mutant, indicating that cysD is positively regulated by c-di-GMP and BrpT. Thus, not only does the production of PAPS trigger brp expression, but the activity of BrpT, the transcriptional activator of brp expression, biases the sulfur assimilation pathway to favor the accumulation of PAPS (i.e., the expression of genes directing PAPS production is increased while the expression of genes that would deplete the PAPS pool is repressed).

DISCUSSION
c-di-GMP regulates the expression of the brp-encoded extracellular polysaccharide that promotes V. vulnificus biofilm formation and rugose colony development (7). Temperature is one of the few primary environmental signals known to affect brp expression in V. vulnificus (12). Here, we identify calcium as a primary signal regulating biofilm formation in V. vulnificus. Calcium, second to magnesium in divalent abundance in seawater (28), specifically triggered brp expression. Genetic screening revealed that  an intermediate of the sulfate assimilation pathway (PAPS) was part of the regulatory circuitry altering brp expression in response to changes in environmental calcium concentration. Our genetic data suggested that genes required for the production of PAPS in response to extracellular calcium affected brp expression in V. vulnificus. Mutation of cysD, cysC, or cysH is known to alter the intracellular level of the PAPS intermediate in other bacteria (29)(30)(31)(32)(33). We showed that cysD and cysC null mutants, which cannot produce PAPS, could not induce P brpA in response to increased environmental calcium, whereas a ΔcysH mutant, which accumulates PAPS, exhibited increased P brpA expression even in the absence of added calcium. Although we did not directly quantify intracellular PAPS levels (the modified nucleotide is highly unstable) (34), intracellular c-di-GMP (and presumably PAPS) levels are elevated in a V. vulnificus cysH mutant relative to the wild type. Thus, mutations in genes that decreased (cysD and cysC) or increased (cysH) PAPS production influenced calcium-induced brp expression. Moreover, elevation of intracellular c-di-GMP levels increased the expression of genes required for PAPS production (cysDNC), and this increase was BrpT dependent. c-di-GMP also repressed the expression of cysHIJ, the products of which drain the PAPS pool. Thus, this feed-forward signaling network biased the sulfur metabolism pathway toward PAPS accumulation, which in turn increased cellular c-di-GMP levels, which increased brpT expression, which ultimately increased production of both Brp polysaccharide and PAPS. Disruption of brpF or brpJ (which encode a glycosyltransferase and a transport protein, respectively) did not affect calcium-induced brp expression, suggesting little or no feedback inhibition of P brpA expression due to incomplete polysaccharide biosynthesis or failed export of a completed polysaccharide. It was recently reported that lesions in cysH, cysJ, cysK, and cysN negatively impacted rugose colony development in Vibrio fischeri (35). However, the defect was most severe in the cysK mutant, while disruption of cysN had little impact, suggesting that the mechanism must differ from that in V. vulnificus.
A second messenger is an intracellular signaling molecule whose concentration increases or decreases in response to the binding of an extracellular ligand to a cell surface receptor. Thus, PAPS behaves as a second messenger in V. vulnificus, akin to c-di-GMP. We propose that elevated environmental calcium increases PAPS production, which ultimately mediates brp expression by altering the cellular c-di-GMP pool. PAPS production may regulate eps production in Vibrio parahaemolyticus as well; lesions in cysD and cysC prevented pellicle formation at the air-liquid interface, a phenotype that is dependent on expression of the cps polysaccharide locus (36). A V. vulnificus cysD mutant exhibited the same defect. This is noteworthy since V. vulnificus and V. parahaemolyticus occupy similar environmental niches (37). Our data suggested that PAPS levels also affected the production of a biofilm matrix component, in an addition to the Brp polysaccharide, that is essential for rugosity. A candidate is the calcium-binding matrix protein CabA, which is required for biofilm and rugose colony formation (17). It was recently demonstrated that cabABC expression in V. vulnificus is c-di-GMP and BrpT dependent (16). Although elevation of the cellular c-di-GMP level (conditions under which CabA should be produced) overcame the calcium-induced P brpA expression defect of the V. vulnificus cysD and cysC mutants, it was not sufficient to support rugose colony development. PAPS may influence CabA activity, or PAPS may regulate the production of another unidentified matrix component.
APS and PAPS are modified nucleotide intermediates of sulfate assimilation that have both been shown to have regulatory effects on gene expression. In Escherichia coli, APS acts as a signaling molecule that is directly bound by the transcriptional regulator Cbl. This results in the inhibition of genes required for organosulfonate utilization and thus helps to establish the hierarchy of preferred sources of sulfur (38). Expression of the csgBAC locus that governs curli production in E. coli is controlled by the regulator CsgD (39), and csgD expression is regulated by the PDE YciR (40). Hence, csgBAC expression is c-di-GMP dependent. It was recently demonstrated that csgBAC expression increased in a ΔcysH strain, while a ΔcysD ΔcysH double mutant or overexpression of cysH decreased csgBAC transcription (33). Since, csgBAC expression is Calcium Triggers Biofilm Formation in V. vulnificus ® affected by the intracellular concentration of both c-di-GMP and PAPS, and csgBAC expression increased on a ΔcysH strain, this suggests that PAPS accumulation must lead to altered cellular c-di-GMP concentrations and suggests that the signaling link we identified between the two second messengers in V. vulnificus may be evolutionarily conserved in E. coli and potentially other Gram-negative bacteria. Ultimately, PAPS appears to mediate the expression of genes for cell surface modification in response to environmental stimuli (curli in E. coli and EPS in V. vulnificus and V. parahaemolyticus). Our study clearly demonstrates interplay between PAPS and c-di-GMP signaling in V. vulnificus and PAPS was recently shown to impact cAMP receptor protein (CRP)-cAMP regulatory networks in E. coli (41), suggesting overlap in the pathways regulated by these three modified nucleotides (cAMP, c-di-GMP, and PAPS). Collectively, these studies highlight the widening regulatory influence of PAPS on bacterial physiology and its intersection with other modified nucleotide-regulated pathways positions PAPS as a potential global signaling molecule.
We propose the following model for the integration of environmental calcium signaling and PAPS production in the regulation of brp expression (Fig. 8). Environmental calcium upregulates sulfate assimilation. Sulfate that is taken up is converted to APS by the action of CysD and CysN and converted to PAPS by CysC. The accumulation of PAPS triggers an increase in the intracellular c-di-GMP level, which in turn increases brpT expression via the regulator BrpR, while simultaneously repressing expression of BrpT and c-di-GMP bias the sulfate assimilation pathway toward the accumulation of PAPS in a feed-forward regulatory network (framed in green). Points of activation and repression are indicated by solid blue arrowheads and flat lines, respectively. Environmental calcium stimulates sulfate uptake. The sulfate is converted to adenosine 5=-phosphosulfate (APS) by the action of CysD and CysN and converted to 3=-phosphoadenosine 5=-phosphosulfate (PAPS) by CysC. The accumulation of PAPS triggers an increase in the intracellular c-di-GMP level, which increases brpT expression via the regulator BrpR and also represses expression of the cysJIH operon that drains the PAPS pool by processing it to PAP and sulfite and, eventually, to cysteine. BrpT activates expression of the brp exopolysaccharide, cab matrix genes, and cysDNTC operons (blue, purple, and green block arrows, respectively), which generates more PAPS, c-di-GMP, and BrpT, thus establishing a feed-forward signaling network that biases the sulfur metabolism pathway toward accumulating PAPS to increase brp expression and biofilm formation. PAPS also has a separate and distinct impact on rugose colony development. The operon structure of cysDNTC and cysJIH (red block arrow) allows for their simultaneous regulation by c-di-GMP and, for the regulation of cysDNTC, by BrpT. Environmental calcium may also influence cellular c-di-GMP levels independent of sulfate assimilation (dashed blue line). Other environmental cues (e.g., temperature) can directly affect intracellular c-di-GMP concentrations that may then feed into the PAPS-c-di-GMP loop.
the cysJIH operon that normally drains the PAPS pool by processing it into sulfide and eventually cysteine. BrpT activates expression of the brp exopolysaccharide and cys-DNTC operons, which generates more PAPS, c-di-GMP, and BrpT. Thus, a feed-forward regulatory network is established in which calcium, c-di-GMP, and BrpT bias the sulfate assimilation pathway toward the accumulation of PAPS and increased brp expression.
Oysters and other bivalves have the capacity to locally concentrate calcium for shell growth and repair (42). Conceptually, the ability of V. vulnificus to sense a change in environmental calcium concentration and initiate biofilm formation at the transcriptional level makes it a perfect cue to promote niche (oyster) colonization by this pathogen. Calcium can repress Vibrio polysaccharide (VPS)-dependent biofilm formation by V. cholerae (43), although V. cholerae can also form distinct Ca 2ϩ -dependent/ VPS-independent biofilms in which calcium is proposed to interact with negatively charged groups in the O antigen or capsule and physically mediate surface adhesion (44,45). The different physiological biofilm responses of V. vulnificus and V. cholerae to calcium under certain conditions may help explain how these closely related species use similar biofilm networks yet evolved to occupy distinct niches in a shared environment. The adoption of calcium as a primary signaling molecule for V. vulnificus biofilm formation is ideal, and it is intriguing to think that V. vulnificus may have in fact evolved to exploit calcium as a molecular beacon to indicate preferred environments for colonization.
Fluorescent reporter fusion analysis. The wild-type strain bearing the P brpA gfp reporter was grown overnight, diluted 100-fold in fresh medium containing the indicated salts, and inoculated in triplicate into a 96-well plate. After 16 h at 30°C, the optical density at 600 nm (OD 600 ) and florescence measurements were read on a BioTek Synergy H1 microplate reader (excitation, 485 nm; emission, 528 nm). Data of three technical replicates from three biological replicates were collected for all experiments. Plots show the mean, and error bars represent the standard deviation.
␤-Galactosidase assays. Quantification of ␤-galactosidase activity was carried out following the original Miller protocol (46) with the following amendments. Briefly, 1 OD 600 unit of cells was pelleted by centrifugation for 10 min at 12,000 ϫ g. Cell pellets were resuspended in 1 ml Z buffer (0.06 M Na 2 HPO 4 , 0.04 M NaH 2 PO 4 , 0.01 M KCl, 0.001 M MgSO 4 , 0.05 M ␤-mercaptoethanol, pH 7.0). Next, 50 l on 1% SDS was added and the mixture was vortexed. Chloroform (100 l) was added, and the sample was vortexed for 10 s and then centrifuged for 2 min at 12,000 ϫ g. A 150-l aliquot of each sample was transferred to the well of a 96-well plate containing 50 l (10 mg/ml) of o-nitrophenyl-␤-D-galactopyranoside (ONPG) and placed in a BioTek Synergy HT microplate reader, where A 420 and A 595 measurements were taken in 5-min intervals. Modified Miller units were calculated using the formula described in the original assay. All assays were performed in triplicate.
Tn mutagenesis. The mini-Tn10 delivery vector pNKTXI-SceI (47) was used for transposon mutagenesis. Cultures of V. vulnificus bearing a P brpA lacZ reporter and E. coli bearing our Tn on a replicating plasmid were grown at 30 and 37°C, respectively. After overnight growth, cultures were diluted by 1:100 and grown to an OD 600 of 0.5. One milliliter of each culture was washed of antibiotics by pelleting with centrifugation and resuspended in 100 l of LB. The E. coli-V. vulnificus mixture was then placed on a nitrocellulose filter disc on an LB plate and placed at 37°C for conjugation. After overnight growth, cells from the filter disc were resuspended in 5 ml LB and plated onto LB with rifampin, 120 g/ml X-Gal, 2 g/ml chloramphenicol, 160 g/ml kanamycin, and 15 mM CaCl 2 . After 48 h of growth, white colonies (representing loss of brp expression in the presence of calcium) were rescreened on plates containing a 10 mM mixture of rifampin, X-Gal, chloramphenicol, kanamycin, and CaCl 2 . A total of 30,000 mutants were screened on LB Ca plates containing X-Gal for isolates exhibiting decreased P brpA lacZ expression. Identified mutants were pooled, and collective genomic DNA was prepared and subjected to wholegenome sequencing at the Indiana University-Bloomington Center for Genomics and Bioinformatics (IUB-CGB). Geneious (Biomatters) was used to map the reads to the mini-Tn10 reference sequence and recover flanking V. vulnificus genomic sequences that were mapped back against the whole genome of strain ATCC 27562 (19). In-frame markerless deletions of target genes were generated by splicing by overhang extension (SOE) PCR (see below).
Mutant construction. Allelic replacement of the target gene was achieved by SOE PCR as follows. PCR fragments corresponding to 3 kb upstream and downstream of the target gene (for primers, see Table S2 in the supplemental material) were amplified and stitched to a central trimethoprim or Calcium Triggers Biofilm Formation in V. vulnificus ® kanamycin antibiotic resistance cassette. Recipient strains carrying the pMMB-TfoX expression plasmid (48) were induced overnight in LB containing ampicillin and IPTG at 30°C. A 10-l aliquot of cells was added to 500 l of Instant Ocean at 20 ppt (IO-20) containing IPTG and 25 l of the SOE PCR product. The transformation mixture was incubated statically overnight at 30°C. The next day, 1 ml of LB was added to the tube, and the cells were allowed to outgrow for 6 h before being plated on LB with selective antibiotics. Allelic replacement of the target gene was confirmed by PCR.
Rugose colony morphology assays and crystal violet staining of biofilms. For visualization of P brpA lacZ expression, strains were grown overnight on LB plates that included the appropriate antibiotics, supplements, and X-Gal at 30°C. Images were captured after 48 h using a Leica MS5 dissecting scope equipped with a Leica DC300F camera. To assess rugose colony morphology, 2-l aliquots of overnight cultures were spotted onto the indicated LB plates and incubated at 30°C for overnight, followed by an additional day at room temperature. Colony morphology images were captured as described above. A fixed radius was used to determine the surface area for colonies with the SP package (49) in R (50). Corresponding 3D elevation maps were created with ImageJ (51). A representative image of samples done in triplicate is shown.
For crystal violet staining, plastic coverslips were placed into 12-well tissue culture plates containing 3 ml of LB medium plus amendments and inoculated with an exponentially growing culture of the indicated strain to an OD 600 of 0.025. After 16 h of static growth at 30°C, coverslips were removed and placed into 4 ml of 0.1% crystal violet for 10 min to stain. The coverslips were rinsed 3 times in 4 ml VPBS (130 mM NaCl, 5 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , pH 7.4 [52]), and the violet stain was solubilized in 4 ml of 4:1 isopropanol-acetone. The A 595 reading was normalized to the OD 600 for each sample. All assays were done in triplicate.
qRT-PCR. The primers used to monitor expression of the indicated genes are listed in Table S2. Strains carrying an arabinose-inducible plasmid expressing the DGC dcpA (7) were grown with antibiotics overnight at 30°C. The next morning, an aliquot was diluted 1:100 into fresh medium and grown to an OD 600 of 0.5. The culture was then diluted into 5 ml of LB containing 0.1% arabinose (to induce DcpA expression) to an OD 600 of 0.025 and grown for 5 h at 30°C. RNA was extracted from 1 OD unit of pelleted cells using the Bio Basic Total RNA miniprep kit. qRT-PCR was carried out using the SensiFAST SYBR Hi-ROX one-step kit. Reactions were performed on an Applied Biosystems StepOnePlus real-time PCR system. Data were analyzed using double threshold cycle (ΔC T ) analysis. All assays were done in triplicate.
Monitoring biofilm development in microfluidic chambers. Polydimethylsiloxane (PDMS)-glass flow cell devices containing eight 40-by 5-by 1-mm chambers were fabricated and sterilized (53). Mid-log-phase wild-type cells (OD 600 of 0.1) were seeded into separate flow cell chambers. Control chambers for background fluorescence contained medium only. Initial attachment (no flow) proceeded for 20 min, followed by a flow rate of 0.25 ml min Ϫ1 in LB or LB supplemented with 5 or 10 mM CaCl 2 for 16 h. The chambers were then flooded with the same medium containing 2.5 M Syto21 (Molecular Probes) for 10 min. Biofilm images and z-stacks (20 ϫ 1-m slices) were captured with an Olympus IX83 microscope using a UPLSAPO 40ϫ silicon oil immersion objective (NA, 1.25; WD, 0.3 mm). Background fluorescence was subtracted from each sample, and quantitative analysis to determine biomass was performed using cellSense (Olympus) and Comstat (54). Data from three biological replicates were analyzed for each strain. The images presented are from a single representative experiment.
Quantification of the intracellular c-di-GMP concentration. c-di-GMP was extracted from whole cells as previously described (12). Briefly, the indicated strains were grown overnight in LB or LB Ca with antibiotics when appropriate. The next day, cells were diluted 1:100 in fresh medium and grown to an OD 600 of 0.5. Cells were again diluted to an OD 600 of 0.025 into fresh LB supplemented with the indicated cations and L-Ara where indicated. Cells were grown for 12 h, when peak brp expression was observed in the presence of CaCl 2 . Six OD units of each sample was centrifuged at 4°C for 10 min at 12,000 ϫ g. The pellet was resuspended in 200 l of 40% (vol/vol) acetonitrile-40% (vol/vol) methanol-20% 0.1 N formic acid. This suspension was incubated at Ϫ20°C for 30 min and then centrifuged at 4°C for 5 min. The supernatant was recovered, and 10 l of each sample was analyzed by liquid chromatographytandem mass spectrometry (LC-MS/MS) on a Quattro Premier XE mass spectrometer (Waters) coupled with an Acquity Ultra Performance LC system (Waters). c-di-GMP was detected by electrospray ionization using multiple reaction monitoring in negative-ion mode at m/z 689.16¡344.31. The MS parameters were as follows: capillary voltage, 3.5 kV; cone voltage, 50 V; collision energy, 34 V; source temperature, 110°C; desolvation temperature, 350°C; cone gas flow (nitrogen), 50 liters/h; desolvation gas flow (nitrogen), 800 liters/h; collision gas flow (nitrogen), 0.15 ml/min; and multiplier voltage, 650 V. Chromatography separation was reverse phase using a Waters BEH C 18 2.1-by 50-mm column at a flow rate of 0.3 ml/min with a gradient of 10 mM tributylamine plus 15 mM acetic acid in 97:3 water-methanol (solution A) to methanol (solution B) with the following parameters: t ϭ 0 min; 99% A:1% B, t ϭ 2.5 min; 80% A:20% B, t ϭ 7.0 min; 35% A:65% B, t ϭ 7.5 min; 5% A:95% B, t ϭ 9.01 min; and 99% A:1% B, t ϭ 10 min. A c-di-GMP standard curve for calculating the c-di-GMP concentration in each extract was generated by dissolving synthesized c-di-GMP (Axxora) in extraction buffer to 250 nM, followed by 2-fold serial dilutions to 0.975 nM. Data from three biological replicates were analyzed for each strain.
Statistical analyses. Statistical significance was determined by the Student t test (two-tailed distribution with two-sample, equal variance) when directly comparing two conditions or a one-way analysis of variance (ANOVA) followed by pairwise comparisons with a Bonferroni adjustment when comparing data with multiple samples.