This Article Abstract Full Text (PDF) Other Versions of this Article: JB.01510-06v1 189/11/4094    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kim, Y.-K. Articles by McCarter, L. L. Search for Related Content PubMed PubMed Citation Articles by Kim, Y.-K. Articles by McCarter, L. L.

ScrG, a GGDEF-EAL Protein, Participates in Regulating Swarming and Sticking in Vibrio parahaemolyticus

Department of Microbiology, The University of Iowa, Iowa City, Iowa 52242

Received 26 September 2006/ Accepted 18 March 2007

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this work, we describe a new gene controlling lateral flagellar gene expression. The gene encodes ScrG, a protein containing GGDEF and EAL domains. This is the second GGDEF-EAL-encoding locus determined to be involved in the regulation of swarming: the first was previously characterized and named scrABC (for "swarming and capsular polysaccharide regulation"). GGDEF and EAL domain-containing proteins participate in the synthesis and degradation of the nucleotide signal cyclic di-GMP (c-di-GMP) in many bacteria. Overexpression of scrG was sufficient to induce lateral flagellar gene expression in liquid, decrease biofilm formation, decrease cps gene expression, and suppress the scrABC phenotype. Removal of its EAL domain reversed ScrG activity, converting ScrG to an inhibitor of swarming and activator of cps expression. Overexpression of scrG decreased the intensity of a ³²P-labeled nucleotide spot comigrating with c-di-GMP standard, whereas overexpression of scrG_EAL enhanced the intensity of the spot. Mutants with defects in scrG showed altered swarming and lateral flagellin production and colony morphology (but not swimming motility); furthermore, mutation of two GGDEF-EAL-encoding loci (scrG and scrABC) produced cumulative effects on swarming, lateral flagellar gene expression, lateral flagellin production and colony morphology. Mutant analysis supports the assignment of the primary in vivo activity of ScrG to acting as a phosphodiesterase. The data are consistent with a model in which multiple GGDEF-EAL proteins can influence the cellular nucleotide pool: a low concentration of c-di-GMP favors surface mobility, whereas high levels of this nucleotide promote a more adhesive Vibrio parahaemolyticus cell type.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The marine bacterium and pathogen Vibrio parahaemolyticus senses and adapts to communal life on surfaces (reviewed in reference 38). On contact with surfaces, it can differentiate to a swarmer cell, which is a specialized cell type suited to group migration over surfaces and through viscous environments. It can also grow in robust biofilm communities and form cohesive pellicles at air-liquid interfaces (18). Prior genetic analysis has identified a locus controlling two gene systems that are pertinent to growth on surfaces (9). The locus, named scr, for "swarming and capsular polysaccharide regulation," encodes the three-gene operon scrABC. Overexpression of scrABC positively regulates lateral flagellar gene expression and negatively regulates capsular polysaccharide (CPS) gene expression; however, the products of the operon do not encode transcription factors. Rather, the operon encodes a potential pyridoxal-phosphate-dependent enzyme, an extracellular solute binding protein, and a membrane-bound GGDEF-EAL motif sensory protein.

GGDEF and EAL domains, which are named after conserved signature amino acids, are highly conserved protein domains that are found throughout the bacterial kingdom (reviewed in references 20, 22, and 53). Biochemical roles for these domains were originally discovered and elucidated in Gluconacetobacter xylinus (formerly Acetobacter xylinum) (55, 56). Six purified proteins containing both GGDEF and EAL domains were demonstrated to possess either diguanylate cyclase or phosphodiesterase activity (64). The opposing activities of these enzymes help determine the intracellular concentration of the nucleotide cyclic di-GMP (bis-3',5'-cyclic dimeric GMP [c-di-GMP]), which serves as a specific allosteric activator of cellulose synthesis in G. xylinus. This role for c-di-GMP in activating cellulose biosynthesis has now been extended to many other bacteria. GGDEF proteins cloned from various bacteria can elicit cellulose production in Rhizobium and Agrobacterium species (2, 4). In addition, GGDEF proteins have been shown to induce cellulose production in other bacteria, such as Salmonella enterica serovar Typhimurium, Salmonella enteritidis, Escherichia coli, and Pseudomonas species (14, 23, 51, 54, 63, 68, 72).

Proteins with GGDEF domains (160 amino acids [aa]; COG2199) constitute one of the largest clusters of potential orthologues in bacteria. The EAL domain (240 aa: COG2200) is similarly ubiquitous (20). V. parahaemolyticus is notably well endowed, with 28 GGDEF-containing proteins, 12 EAL-containing proteins, and 15 proteins with both domains (netting 55 GGDEF- and/or EAL-containing proteins). Additionally, there are other proteins with potential phosphodiesterase activity that may contribute to modulation of the c-di-GMP pool, i.e., proteins with HD-GYP domains (COG2206) (19). V. parahaemolyticus contains at least two such proteins. Bioinformatic analysis first suggested that the GGDEF domain, which shows homology to the adenylyl cyclase catalytic domain, possesses diguanylate cyclase synthetic activity (47). This prediction has been experimentally supported by the ability of proteins solely containing the GGDEF domain to produce c-di-GMP (28, 46, 58, 62). Phosphodiesterase activity has been demonstrated to be associated with the EAL domain (6, 13, 29, 60, 65). Purified composite proteins with both domains—as the six enzymes from G. xylinus—generally display either diguanylate cyclase or phosphodiesterase activity (61). This has resulted in the hypothesis that only one domain is catalytically active, whereas the second domain may be nonfunctional or regulatory (12, 56).

The majority of GGDEF and EAL domains are found in combination with a variety of signal reception/transduction domains; thus, there seems an immense potential for the activity of GGDEF-EAL proteins to be regulated by diverse signals (20-22). For example, GGDEF was first identified as a novel signaling domain in the Caulobacter crescentus response regulator PleD, which participates in polar-to-stalked-cell development (27). Phosphorylation of the receiver domain in PleD leads to polar localization of PleD, activation of its GGDEF output domain, and production of c-di-GMP (46). Furthermore, mutational analyses demonstrated that the highly conserved signature residues (actually GGEEF in PleD) are essential for output function controlling the swarmer-to-stalked-cell transition (1). A different signal transduction module, a PAS domain, is found linked to the GGDEF and EAL domains of phosphodiesterase A1 of G. xylinus; its c-di-GMP-degradative activity has been shown to be regulated in response to oxygen (11). V. parahaemolyticus ScrC is a membrane protein that contains a potential periplasm-sensing domain, and ScrG (the GGDEF-EAL protein studied in this work) contains a PAS domain.

c-di-GMP clearly influences many more cellular processes than cellulose biosynthesis: a signaling role for c-di-GMP, acting as a second messenger, to influence the production of a variety of cell surface determinants is now being found in many bacteria (reviewed in references 16, 31, and 52). One of the best-studied examples is the involvement of c-di-GMP in promoting the multicellular rdar morphotype of S. enterica serovar Typhimurium (reviewed in reference 50). Production or degradation of c-di-GMP by proteins containing GGDEF, EAL, or HD-GYP domains has now been demonstrated in many organisms, including Pseudomonas and Xanthomonas spp., Shewanella oneidensis, Vibrio cholerae, Vibrio fischeri, and Yersinia pestis (6, 28, 33, 36, 45, 57, 58, 62, 66, 67). GGDEF-EAL signaling seems to be particularly pertinent in the context of surface colonization and biofilms. In addition to cellulose, GGDEF-EAL proteins control the production of other constituents of the extracellular matrix, including CPS, cell surface proteins, flagella, curli, and pili (5, 9, 10, 15, 24, 25, 28, 30, 34, 49, 50, 62).

In this work, we identify and characterize the gene scrG, which participates in the control of swarming and CPS production. ScrG contains both GGDEF and EAL conserved domains, although the GGDEF domain is rather poorly conserved and lacks the signature GGDEF motif. The gene was discovered in a screen for low-copy-number cosmid clones that induced expression of lateral flagellar genes in liquid, a condition normally repressing for laf gene expression. The scrG gene was subcloned from the cosmid and shown to be solely responsible for the laf-activating ability. A mutant strain (scrG::Tn5) was constructed and found to have alterations in swarming motility and colony morphology. Together, the genetic evidence and analysis of cellular nucleotide pools presented in this work suggest that ScrG acts as a phosphodiesterase. In V. parahaemolyticus, c-di-GMP-mediated regulation serves to coordinate the balance between the cell's capacities to stick and to be mobile by reciprocally regulating transcription of capsular polysaccharide (cps) and lateral flagellar (laf) gene expression.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bacterial strains and growth conditions. The strains and plasmids used in this work are described in Table 1. All V. parahaemolyticus strains are descendants of the translucent strain BB22TR (19) and were routinely grown at 30°C. Heart infusion medium (heart infusion broth) contained 25 g of heart infusion (Difco) and 15 g of NaCl per liter. Heart infusion swarm plates were prepared by adding 15 g/liter Bacto agar (Difco) to heart infusion broth. Heart infusion swarm minus plates were prepared with 20 g/liter granulated agar. Difco Bacto agar is more conducive for swarming than is Difco granulated agar. Heart infusion plates were also amended with different amounts of NaCl, FeCl₃, and/or agar as indicated to make plates less permissive for swarming. Marine LB plates (LMB) contained 10 g of tryptone, 15 g of NaCl, 5 g of yeast extract, and 20 g of granulated agar per liter. Congo red plates contained 25 g of heart infusion and 20 g of granulated agar per liter, to which 10 ml of 2.5% Congo red (in ethanol) and 5 ml of 1 M CaCl₂ were added after autoclaving. Swimming motility plates contained 10 g of tryptone, 15 g of NaCl, and 3 g of Bacto agar per liter. X-Gal (5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside) was dissolved in N',N-dimethyl formamide and used at 75 µg/ml. Antibiotics were used at final concentrations of 50 µg/ml kanamycin, 10 µg/ml chloramphenicol, 25 µg/ml gentamicin, and 10 µg/ml tetracycline, unless otherwise indicated. Overexpression plasmids were induced with 1 mM isopropyl-ß-D-galactopyranoside (IPTG).

Cosmids affecting lateral flagellar gene expression were identified by using a V. parahaemolyticus library, as described previously (9). Briefly, a cosmid library was conjugated into the laf::lux reporter strain, and transconjugants were screened for the ability to produce light in 200 µl of heart infusion broth (plus tetracycline) in microtiter wells. The library was made with the vector pLAFRII, and each clone in the library contained 35-kb insertions of wild-type V. parahaemolyticus DNA (42). One of these luminescence-inducing cosmids was pLM2550, the scrG-containing clone studied in this work. Transposon mutagenesis using the Tn5-derived transposon Tn5tac1 (12) was performed on the V. parahaemolyticus tetracycline-resistant cosmid pLM2550 in E. coli. Mutagenized cosmids bearing the kanamycin resistance transposon were transferred to the V. parahaemolyticus laf::lux reporter strain LM1017, and the transconjugants were screened for loss of light production in heart infusion (plus tetracycline) in microtiter wells. The point of insertion of the transposon on the mutant cosmid pLM3185 was identified by DNA sequencing to occur within the codon for aa 460 of ScrG.

The scrG gene was subcloned by isolating a 2.2-kb NdeI-HpaI fragment from pLM2550 and ligating the fragment into the SmaI site of the IPTG-inducible expression vector pLM1877. scrG3 removed DNA coding for aa 9 to 556; scrG2 (scrG_EAL) removed the EAL domain encoded between aa 309 and 556. The deletions were replaced with a chloramphenicol resistance cassette. These deletion/insertion mutations were made by using a Red recombinase system in E. coli (17) to introduce Cam^r cassettes onto cosmid pLM2550 or onto the expression plasmids pLM3176, pLM3531, and derivatives.

Site-directed mutations were introduced by a single-step procedure for mutagenizing large plasmids (69). The scrG subclone in vector pLM1877, plasmid pLM3176, proved too large for a successful yield of the PCR product, and so scrG was subcloned into the smaller IPTG-inducible expression vector pDSW361 to make pLM3531. pDSW361 is a kanamycin-resistant derivative of pDSW204 (70). High-performance liquid chromatography-purified mutagenic primers were 50 nucleotides in length. DNA sequencing of the entire scrG gene confirmed successful introduction of the specific mutation. Kanamycin-resistant pDSW361 and derivatives were transferred to V. parahaemolyticus by electroporation.

Surface and liquid luminescence assays. For time course experiments on plate-grown cells, strains were pregrown on plates and suspended to an optical density at 600 nm (OD₆₀₀) of 0.01, and 100 µl was spread onto multiple fresh heart infusion swarm plates (with antibiotics and IPTG when appropriate). Plates were suspended at the specified times in 5 ml of heart infusion broth, and OD and light unit measurements were recorded. Several plates were harvested at early time points to ensure adequate cell numbers. For time courses in liquid, overnight heart infusion cultures were diluted to an OD₆₀₀ of 0.05 into 250-ml flasks containing 25 ml of heart infusion broth (and IPTG when appropriate). These cultures were incubated with aeration at 30°C, with periodic sampling for OD₆₀₀ and light readings. Luminescence was quantified by measuring 0.1-ml samples for 30 s in a Turner TD20/22 luminometer (Turner Designs). Dilutions were made to keep all measurements within a linear range. Luminescence is reported as specific light units (SLU), which are total light units per minute per milliliter per OD₆₀₀ unit. Light readings were taken in triplicate and standard deviations were generally less than 10%. Each experiment was performed at least two times.

ß-Galactosidase measurements. Cells were grown as described above for luminescence measurements. Assays for LacZ reporter activity were performed according to Miller (44). Cells were permeabilized with Koch's lysis solution (48). Each experiment was performed at least two times, with measurements for each sample performed in triplicate. Standard deviations were generally less than 10%.

Biofilm quantification. To quantify pellicle formation and adherence to glass, 2-ml heart infusion broth cultures were grown overnight at 30°C with shaking. Replicate cultures (minimum of six per experiment) were subsequently decanted, and the residual cells adhering to the wall of the test tube were suspended by vortexing in 2 ml of heart infusion broth for measurements of OD₆₀₀. Experiments were repeated at least three times, and a representative experiment, with standard deviation, is presented.

Immunoblot analysis. Detection of lateral flagellin was performed as described previously (8). Usually, samples were diluted to a concentration of 0.025 OD₆₀₀ units, and 5 µl was loaded per well. Pooled antisera to both polar and lateral flagellins were used at dilutions of 1:10,00 and 1:2,500, respectively. The antisera also cross-reacted with a nonflagellar protein that served as a normalization control. Quantification was performed by using the ImageGauge program, version 3.12 (Fuji Photo Film Co., Ltd.); averages were calculated from at least three immunoblots, each with independently grown samples.

Detection of c-di-GMP. Labeling and analysis of cellular nucleotides were performed essentially as described by Bochner and Ames (7). Starter cultures were grown overnight in MOPS (morpholinepropanesulfonic acid) minimal medium (7) with 150 mM NaCl, 25 mM MgCl₂, 0.4% galactose, 50 µg/ml gentamicin, and 1 mM IPTG. This medium contained no added phosphate but was supplemented with 0.07% Casamino Acids. Cells were diluted to an OD₆₀₀ of 0.02 into fresh medium also containing ³²P_i and grown at 34°C (usually 12 h). Carrier-free ³²P in acid-free aqueous solution was purchased from Amersham Biosciences and used at 100 µCi/ml of medium. Aliquots were removed at the indicated times, mixed with 0.1 volume 11 N cold formic acid, and placed on ice. Extracts were applied to prewashed (in 0.5 M LiCl) PEI (polyethyleneimine) cellulose plates (J. T. Baker Chemical Co.). After spotting of the extract, plates were soaked in methanol and air dried. Thin-layer chromatography (TLC) solvents were adopted from previously published procedures (55, 67). Development in the first dimension was performed in 0.2 M NH₄HCO₃, usually for 120 min. After a wash in methanol, chromatography in the second dimension (150 min) was performed in 1.5 M KH₂PO₄, pH 3.65 (adjusted with o-phosphoric acid). Plates were washed in methanol after chromatography and prior to autoradiography. Exposure to X-ray film was usually for 1 to 3 days. Two sources of standards were used for c-di-GMP: one was generously provided by the laboratory of M. Benziman, and the second was purchased from Biolog (Bremen, Germany). It should be noted that the c-di-GMP spot continued to migrate after the solvent front reached the top of the plate, particularly in the first dimension.

ODs of parallel cold cultures were recorded at the time of formic acid extraction of the labeled cultures, and the amount of extract loaded on the chromatogram was normalized to a final OD₆₀₀ (usually 2 to 4 µl/sample). Quantification of spot intensity was performed with the image quantification program of Molecular Dynamics (ImageQuant 5.2). For each autoradiogram in a series (i.e., vector, ScrG, and ScrG_EAL labeled in the same experiment), the spot intensity for the c-di-GMP spot was divided by the intensity of spot O. Spot O, which migrates directly above c-di-GMP, was chosen as the reference spot because its intensity was similar in magnitude to the c-di-GMP spot. (Similar results were obtained when the ATP spot was used as the reference). For each series, the c-di-GMP/spot O ratios were normalized by dividing by the ratio produced for the vector control sample (i.e., the c-di-GMP/spot O ratio was normalized to 1 in the vector sample). Quantification was performed on autoradiograms for three independent labeling experiments (standard deviations of normalized ratios are indicated by error bars).

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Identification of a gene inducing lateral flagellar gene expression in liquid. V. parahaemolyticus induces its lateral flagellar gene system specifically when grown on a surface (41). Using a luminescence reporter transcriptionally fused to a laf gene, we screened a cosmid library for regulatory genes that would inappropriately induce laf expression in liquid culture. Normally, the reporter strain LM1017 (referred to as strain RS313 in reference 39), containing a lux fusion in the lateral flagellar hook operon, produces very little light when grown in broth. Four classes of cosmids, representing distinct regions of the chromosome, were obtained that conferred brilliance upon liquid-grown LM1017. (Additional classes were found conferring lesser degrees of light production.) One of these bright cosmids was characterized previously, and an operon, named scrABC, for "regulation of swarming and capsular polysaccharide," was identified (9). Here, we report the dissection of a second bright regulatory cosmid, pLM2550. Figure 1A shows that this cosmid greatly induced luminescence in liquid, conferring 154,000 SLU, compared to 41 SLU produced by the strain with a control cosmid. Light was monitored throughout the growth cycle, and maximal specific activity (light units per min normalized to OD₆₀₀) is reported; cosmid 2550 consistently conferred enhanced luminescence throughout the growth cycle compared to the vector control.

View larger version (42K): [in this window] [in a new window]   FIG. 1. Identification of scrG. (A) Induction of lateral flagellar (laf) gene expression in liquid. Cultures were grown in heart infusion broth with tetracycline. Strains: LM6377 (with control), LM6378 (with bright cosmid pLM2550), and LM6375 (with dark derivative pLM2550::Tn5). Lateral flagellar gene expression (laf) was measured with the laf::lux luminescence reporter strain LM1017, and light (lux) is reported as SLU. Samples were measured in triplicate, and the experiment shown is representative of three similar experiments. Values shown represent the peak of light production. (B) Colony phenotypes on standard swarm heart infusion (incubated 12 h), modified swarm heart infusion with 3.5 g/liter NaCl and 15 g/liter Bacto agar (incubated for 16 h), and Congo red plates (incubated for 5 days). Colonies of the parental swarm-competent strain (LM5674) are in the top rows, and colonies of the scrG::Tn5 mutant (LM7861) are in the bottom rows of each plate.

The transposon disrupted VP1377 (designated scrG), encoding a 568-aa predicted signal transduction protein. Specifically, this transposon was inserted into the codon for aa 460. The gene does not occur within an operon; flanking open reading frames are transcribed in the opposite orientation to scrG. The domain architecture of the predicted protein product is depicted in Fig. 2. The protein contains three conserved domains: PAS (which has been found to bind a variety of ligands and often acts as sensory domains for light or oxygen), GGDEF, and EAL. The latter two domains are named after highly conserved signature amino acid motifs. Although the E values for the GGDEF domain (5e–16) and the EAL domain (8e–57) are significant, the signature motifs do not exactly match the consensus. Specifically, the reverse-position-specific CD search (37) produced ESL for the EAL signature and HDDDF for the GGDEF motif (Fig. 2B).

View larger version (49K): [in this window] [in a new window]   FIG. 2. Domain architecture: ScrG contains GGDEF and EAL domains. (A) ScrG (568 aa) contains three highly conserved domains: PAS (COG2202; E = 4e–8), GGDEF (COG2199; E = 5e–16), and EAL (COG2200; E = 8e–57). A truncation was engineered in the coding region of scrG to delete aa 309 to 556 and remove the EAL domain (ScrGEAL). Point mutations were also introduced to result in the substitution of alanines for E350, D224, D225, and F226. (B) Conserved domain alignments of ScrG with the EAL and GGDEF domains were performed at http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi.

EAL

View larger version (23K): [in this window] [in a new window]   FIG. 3. Overexpression of scrG is sufficient to induce lateral flagellar gene expression and lateral flagellin production in liquid, and removal of the EAL domain of ScrG alters its activity. (A) laf::lux expression in liquid-grown cultures. (B) Immunoblot of lateral flagellin (Laf) production in liquid-grown cultures. (C) laf::lux expression in strains grown on plates. In the experiments represented in panels A and C, lateral flagellar gene induction was measured in the laf::lux reporter strain LM1017 carrying IPTG-inducible expression plasmids. Strains were grown in heart infusion medium with gentamicin and IPTG. Luminescence is reported as SLU. Luminescence was monitored in triplicate periodically throughout growth of the cultures, and values of maximal expression are shown. Strains: LM6349 (LM1017/vector), LM6353 (LM1017/scrG+), and LM7048 (LM1017/scrGEAL). In the experiments represented in panel B, cultures were grown as in the luminescence experiment (panel A) and harvested for protein analysis when the cells entered stationary phase (6 h after IPTG induction). Immunoblots were loaded with equivalent OD600 units and probed with antisera to lateral (Laf) and polar (Fla) flagellins. The cross-reactive band "X" served as the normalization control. Lanes: 1, plate-grown LM8073 (LM5674/vector); 2, liquid-grown LM8073 (LM5674/vector); 3, liquid-grown LM8074 (LM5674/scrGEAL); 4, liquid-grown LM8075 (LM5674/scrG+).

ScrG (and ScrG_EAL) affects biofilm formation, cps gene expression, and colony morphology. Overexpression of scrG had additional consequences for the cell. V. parahaemolyticus is a proficient biofilm former and produces a copious pellicle at the medium/air interface when cultured in test tubes (18). This pellicle is retained in the test tube when the culture contents are decanted. The top left panel of Fig. 4 shows pellicle formation in strains carrying the vector, scrG, and scrG_EAL expression clones. On the right is quantification of the adherent biofilm. Overexpression of scrG effectively prevented pellicle formation, and the EAL domain was necessary for this activity. Note that all of the experiments represented in Fig. 4 were performed with strains that were deleted for the chromosomal copy of scrG (and similar results were also obtained in the wild-type, scrG⁺, background).

View larger version (63K): [in this window] [in a new window]   FIG. 4. Expression of scrG and scrGEAL affect biofilm formation, cpsA::lac expression, and colony morphology. (A) Biofilm production in test tubes and quantification of the adherent cell population after overnight growth with shaking in heart infusion (plus IPTG and gentamicin). (B) Colony morphology on heart infusion swarm medium (with IPTG, gentamicin, and the chromogenic indicator X-Gal) and quantification of ß-galactosidase activity after 15 h of incubation. Miller units are the averages of activity measured for each of four single colonies. (C) Colony morphology on Congo red plates (with gentamicin and IPTG; incubated for 7 days) and opacity on LMB plates (with gentamicin; incubated overnight). Biofilm and colony morphology effects were examined in LM7863, which contains a deletion of scrG in strain LM5674. ß-Galactosidase activity was examined in LM7005, which contains the scrG deletion and the cpsA::lacZ reporter. Strains (containing the indicated expression plasmids): LM8070 (LM7863/vector), LM8072 (LM7863/scrG+), LM8071 (LM7863/scrGEAL), LM8066 (LM7005/vector), LM8069 (LM7005/scrG+), and LM8068 (LM7005/scrGEAL).

EAL

Expression of scrG and scrG_EAL also affected the appearance of colonies on plates. Colonies containing the scrG clone were more translucent on heart infusion or LMB plates and were smoother on Congo red plates than were the vector-containing colonies. Colonies containing the scrG_EAL plasmid were more opaque on heart infusion, LMB, and Congo red. (LMB plates are easier to photograph than heart infusion plates and are shown in the lower right panel in Fig. 4.) These colony morphology differences were dependent on an intact cpsA locus, as they were not observed in the cpsA::lacZ reporter strain (data not shown). Moreover, these differences in colony morphology could be observed even in the absence of IPTG induction (e.g., in Fig. 4, the LMB medium did not contain IPTG). Therefore, high-level gene expression was not necessary for observation of the colony phenotype.

Examination of cellular nucleotide pools. Proteins with GGDEF and EAL domains are predicted to catalyze the formation or degradation of the nucleotide c-di-GMP. To examine the activity of ScrG, nucleotide pools of cells overexpressing scrG and scrG_EAL were labeled with inorganic ³²P during growth in minimal medium and limiting phosphate, and formic acid extracts were analyzed by two-dimensional TLC (2D-TLC). Figure 5 compares extracts of cells induced with IPTG for scrG and scrG_EAL expression. Three experiments are shown for comparison. In each autoradiogram in Fig. 5, the lower arrow indicates the position corresponding to the migration of the c-di-GMP spot and the upper arrow indicates a reference spot, called spot O. In comparison to the vector control case for each experiment, overexpression of scrG decreased detection of the ³²P-labeled nucleotide spot comigrating with c-di-GMP standard, whereas overexpression of the allele producing a truncated form of the protein lacking the EAL domain but containing the GGDEF domain resulted in enhanced production of this spot. Similar results were also observed with these plasmids in the scrG background (data not shown). The average retardation factor (R_f) for the potential c-di-GMP spot was 0.15 (± 0.02) in the first dimension and 0.31 (± 0.01) in the second dimension; these values are consistent with c-di-GMP standards. They are also consistent with previously reported values for c-di-GMP (0.16 and 0.32 [67] and 0.19 and 0.31 [55]). A reference spot (O) was chosen for normalization: relative to the vector case, overproduction of ScrG caused a decrease in the ratio of the potential c-di-GMP spot compared to O and overproduction of ScrG_EAL caused an increase in the ratio.

View larger version (76K): [in this window] [in a new window]   FIG. 5. Analysis of 32P-labeled nucleotide pools upon IPTG induction of scrG and scrGEAL. Autoradiograms produced from 2D PEI cellulose TLC separations. The solvent used in the first dimension was 0.2 M (NH4)2CO3; the solvent used in the second dimension was 1.5 M KH2PO4, pH 3.65. Development times were 120 and 145 min for the first and second dimensions, respectively. Each row of panels contains an independent labeling experiment of three strains: LM8073 (LM5674/vector), LM8075 (LM5674/scrG+), and LM8074 (LM5674/scrGEAL). In the experiments shown in rows 1 and 2, cells were inoculated at a starting OD600 of 0.02 and labeled for 12 h; in row 3, cells were inoculated at an OD600 of 0.1 and labeled for 7 h. The final OD at the time of harvest was 0.7, and equivalent concentrations were spotted on each TLC plate. The lower arrow in each chromatogram indicates the position of the spot that migrates with Rf values consistent with the mobility of c-di-GMP. The higher arrow indicates the position of spot O, which was used as the normalization control. The ratio of c-di-GMP to spot O was calculated for each autoradiogram, and then the ratio was set to 1 for the vector case and the ratios for the other samples were normalized accordingly.

EAL

View larger version (25K): [in this window] [in a new window]   FIG. 6. Effects of point mutations in the canonical GGDEF and EAL motifs of ScrG. (A) Effects of mutant scrG alleles of on laf::lux expression. Strains with indicated expression plasmids: LM8093 (vector), LM8094 (scrG+), LM8095 (scrGD224A), LM8096 (scrGD225A), LM8239 (scrGF226A), LM8097 (scrGE350A), and LM8140 (scrGEAL). (B) Effects of scrG with EAL mutations on laf::lux expression. Strains with indicated expression plasmids: LM8093 (vector), LM8097 (scrGE350A), and LM8140 (scrGEAL). (C) Effects of GGDEF point mutations in the absence of the EAL domain on laf::lux expression. Strains with indicated expression plasmids: LM8093 (vector), LM8140 (scrGEAL), LM8141 (scrGD224A EAL), LM8142 (scrGD225A EAL), and LM8242 (scrGF226A EAL). All strains were derived from the laf::lux reporter strain LM7004, which contains scrG. Strains were grown in broth culture with IPTG to induce expression of scrG alleles, and samples were taken hourly to monitor OD600 and light production. Luminescence values at time points bracketing maximal light production are shown; however, trends remained consistent throughout the growth curve. Luminescence is reported as SLU. Experiments were repeated at least three times, with similar results.

EAL

The canonical GGDEF sequences are not conserved in ScrG, being HDDDF (Fig. 2B), and so we altered three of these amino acids. Substitution of two of these residues had little effect on the activating ability of ScrG (Fig. 6A). Specifically, expression of scrG with the D224A substitution (HDDDF to HDADF) conferred 500,000 maximal SLU, compared to 1,400,000 SLU conferred by scrG⁺ and 6,000 SLU for the vector. Similarly, the F226A allele (HDDDF to HDDDA) produced effects (1,450,000 SLU) highly similar to the wild-type allele. Mutation of D225A (HDDDF to HDDAF) caused the most perturbation of activity; however, even in this case, the mutant form of the protein still conferred some activation of laf::lux (retaining approximately eightfold activation throughout the growth cycle compared to the vector). Thus, although none of these three residues were required for the potential phosphodiesterase activity, D225 may affect this activity.

Next, we examined whether alteration of the nonconventional GGDEF motif would affect the repressing activity of ScrG_EAL. The three alanine substitution mutations in combination with deletion of the EAL domain had little or no effect on the ability of the truncated proteins to repress laf::lux expression (Fig. 6C).

Relationship between ScrG and ScrC. Effects on colony morphology, CPS production, and swarming suggested that overexpression of scrG_EAL causes an increase in c-di-GMP, and the labeling experiments support such an idea. From prior work, we knew that mutations in the scrABC operon produced a crinkly colony morphology that was the consequence of overproduction of CPS (9). We further knew that the scrABC effects on colony morphology required the transcriptional activator CpsR (26). The V. cholerae homolog of CpsR (VpsR) has also been found implicated in the c-di-GMP control of biofilm formation (67). To investigate whether CpsR was also necessary for the crinkly colony effects observed upon overexpression of scrG_EAL, colony morphology was examined in a cpsR strain. Overexpression of scrG_EAL produced an opaque and crinkly colony in the wild-type background but not in the strain carrying a mutation in cpsR (Fig. 7, part I).

View larger version (38K): [in this window] [in a new window]   FIG. 7. Relationship between ScrG and ScrABC. (I) The scrGEAL expression phenotype requires cpsR. Strains were grown on Congo red plates with IPTG and gentamicin for 5 days. Strains: LM8073 (wild type/vector), LM8074 (wild type/scrGEAL), LM8522 (cpsR/vector), and LM8557 (cpsR/scrGEAL). (II) The scrABC phenotype can be suppressed by expression of scrG, and the EAL domain is necessary for suppression. Strains were grown on heart infusion swarm plates with gentamicin and IPTG. Luminescence is reported as SLU. Luminescence was monitored in triplicate periodically throughout the growth of the cultures, and the time point of maximal expression is shown. Strains: LM7074 (scrABC/vector), LM7075 (scrABC/scrG+), and LM7076 (scrABC/scrGEAL). (III) scrG::Tn5 can be suppressed by scrABC. (A) Swarming was examined on heart infusion plates (containing 5 g/liter NaCl, 15 g/liter Bacto agar, and gentamicin) with or without IPTG, as indicated, after 24 h of incubation. (B) Swarm plates with IPTG were harvested for quantification of lateral flagellin (Laf) production, and the graph shows the normalized average of three independent samples for each strain. The amount of Laf produced by strain 1 (wild type with vector) was set to 100%. (C) Colony morphology on Congo red plates with gentamicin and IPTG after 7 days. Strains: 1, LM 7757; 2, LM8549; 3, LM8550; 4, LM8551; 5, LM8552.

The swarm-defective phenotype of the scrG::Tn5 mutant could be complemented by scrG, the EAL domain was necessary for the complementing activity, and, furthermore, the phenotype could be similarly restored by using an scrABC expression clone (Fig. 7, part III). Extremely high levels of expression in trans were not required, as good swarming was restored even in the absence of IPTG. The effects on lateral flagellin production as well as colony morphology were similar. Complementation and suppression restored lateral flagellin production to wild-type levels (or greater) and reversed the crinkly mutant colony phenotype to restore the smooth, wild-type colony morphology.

Phenotypes of scrG deletion alleles. Overexpression of the various scrG alleles had dramatic consequences with respect to swarming, lateral flagellin production, CPS production (i.e., colony morphology), and gene expression; however, such an analysis could grossly alter nucleotide pools beyond a physiologically pertinent range. Nevertheless, the overexpression experiments clearly revealed that ScrG was capable of contributing to the nucleotide balance with the consequence of affecting swarming and sticking. Moreover, IPTG induction was not necessary for some of these phenotypic effects to be seen, e.g., no IPTG was required to detect changes in opacity on LMB plates in scrG strains with scrG⁺ or scrG_EAL alleles in trans (Fig. 4), suggesting that a high level of expression was not required to influence opacity or swarming. To further probe the contribution of scrG in the regulation of swarming and CPS production, we constructed additional mutants and analyzed their phenotypes.

On heart infusion swarm medium, deletion of scrG or deletion of the EAL domain (scrG_EAL) caused no discernible defect in swarming (Fig. 8), as was also observed for the original Tn5 insertion mutant. When the salt concentration was lowered to slow swarming of the wild type (and the agar concentration balanced appropriately), a small swarming defect was observed for the Tn5 insertion mutant but not for the deletion mutants. However, iron availability is another factor influencing swarming (41), and when iron was added to low-salt swarm plates, the three scrG alleles displayed similarly defective phenotypes with respect to diminished swarming and decreased lateral flagellin production (Fig. 8, part I).

View larger version (43K): [in this window] [in a new window]   FIG. 8. Phenotypes of scrG mutants. (I) Swarming phenotypes of scrG mutants on different kinds of swarming media. Three single colonies of each strain were inoculated in each row. Immunoblot analysis was performed on samples harvested from heart infusion type C plates. Strains: 1, LM5674 (wild type); 2, LM7861 (scrG::Tn5); 3, LM8023 (scrG); 4, LM7863 (scrGEAL). (II) Phenotype of mutants with double scr mutations. Immunoblot analysis was performed on samples harvested from heart infusion type A plates. Strains: 4, LM5674 (wild type); 5, LM5719 (scrC); 6, LM8390 (scrC scrG); 7, LM7863 (scrG). Heart infusion (25 g/liter) plates also contained 15 g/liter NaCl and 15 g/liter Bacto agar (A); 5 g/liter NaCl and 13 g/liter Bacto agar (B); or 5 g/liter NaCl, 13 g/liter Bacto agar, and 50 µM FeCl3 (C). Swim plates contained 10 g/liter tryptone, 15 g/liter NaCl, and 3 g/liter Bacto agar. Plates (from left to right) were incubated for 12, 18, 24, 14, 36, and 8 h. Pooled antisera directed against polar (Fla) and lateral (Laf) flagellins were used. Quantification of Laf production was performed on at least three blots (with independent samples), and the averages with standard deviations are shown in the graphs. In each graph, Laf production was normalized to the control band, indicated as x.

Cumulative effects of mutations in scrG and the scrABC locus were observed to occur at the level of laf transcription. Introduction of the scrG::Tn5 allele into strain LM5545, which carries a polar insertion in scrA and a laf::lux reporter, created a strain (LM6736) that consistently produced less light throughout the growth cycle than did the parental LM5545 strain. For example, maximal light produced by LM5545 (scrA::Cam) was 20,500 ± 1,200 SLU and by LM6736 (scrA::Cam scrG::Tn5) was 1,240 ± 40 SLU. (In this experiment, maximal light production for the wild-type reporter strain LM1017 was 569,000 ± 62,000 and for the scrG::Tn5 single-mutant strain LM6735 was 722,000 ± 23,000.)

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The scrG gene encodes a GGDEF-EAL protein that influences swarming and capsular polysaccharide regulation. This is the second locus encoding GGDEF-EAL proteins discovered to play a role governing the swarming and sticking ability of V. parahaemolyticus. Both loci, scrG and the previously identified scrABC operon, were discovered through screening for low-copy-number cosmids conferring constitutive laf expression (9). Both loci have been found to also influence cell adhesiveness. However, neither GGDEF-EAL protein influences swimming motility. In contrast to ScrG, the GGDEF-EAL protein ScrC is not sufficient by itself to activate lateral flagellar genes, and the other products of the operon (ScrA and ScrB) are required for this activity. Mutants with defects in the scrABC operon show severe swarming defects, and we show here that the phenotype of scrABC mutants can be complemented/suppressed by ectopic expression of scrG and vice versa. Thus, scrG and scrC appear to act in the same regulatory circuit: they can influence expression of the same target genes, their activities can compensate one another, and they both affect colony morphology via the transcriptional regulator CpsR.

ScrG contains three highly conserved domains: PAS, GGDEF, and EAL. The enzymatic activities of many GGDEF and EAL domains have been characterized: the GGDEF domain is sufficient for synthesis of c-di-GMP, and the EAL domain is sufficient for degradation c-di-GMP (reviewed in reference 52). The EAL domain of ScrG is essential for its activity. Expression of a mutant form of the protein carrying a deletion of the entire EAL domain (ScrG_EAL) abolished the ability of ScrG to activate laf in liquid or to complement the scrABC strain. Substitution of an alanine residue for the glutamate (E350A) in the EAL signature also completely abrogated the ability of ScrG to induce laf expression in liquid. The equivalent glutamate residue has also been shown to be essential for the activities of the phosphodiesterases HmsP in Yersinia pestis (6, 34), VieA in V. cholerae (67), and YhjH in S. enterica serovar Typhimurium (62). With respect to cell adhesiveness, overexpression of scrG effectively prevented biofilm formation, and the EAL domain was also critical for this activity. Overexpression of scrG, but not scrG_EAL, in a reporter strain carrying a lacZ insertion in the CPS locus (cpsA::lacZ) caused a decrease in cpsA gene expression compared to a vector control. Thus, because the EAL domain is critical, this genetic evidence suggests that the phosphodiesterase activity of ScrG influences laf and cps gene expression.

Analysis of cellular nucleotide pools by 2D-TLC further supports the assignment of phosphodiesterase activity to ScrG. Expression of scrG but not scrG_EAL decreased the intensity of a ³²P-labeled spot that comigrated with purified c-di-GMP. In fact, the consequence of removal of the EAL domain was that production of this nucleotide spot was enhanced compared to even the vector control.

Does ScrG also have intrinsic diguanylate cyclase activity? Removal of the EAL domain or mutation of the EAL signature did not simply inactivate the protein. Rather, a new activity was observed. ScrG caused activation of lateral flagellar gene expression in liquid. ScrG without its EAL domain prevented laf gene expression on plates. Similarly, with respect to cell adhesiveness, overexpression of scrG_EAL had the reverse effect compared to scrG, causing increased cpsA::lacZ expression and colony opacity and roughness compared to the parent. Clearly, the truncated protein had a new and different activity. The spots in the 2D-TLC analysis were consistent with observed phenotypic effects: the consequence of scrG expression was a less intense spot, whereas scrG_EAL expression resulted in a more intense spot than that of the vector control.

In some respects this observation was surprising. Sequence analysis suggested that ScrG was a likely example of a c-di-GMP-degrading protein with a phantom, enzymatically inactive c-di-GMP-synthesizing domain, because the GGDEF signature was poorly conserved (being HDDDF). To further probe the structure/function of ScrG, site-directed alanine substitution mutations were introduced into the codons for three specific amino acids corresponding to the signature motif by sequence alignment (to make proteins with HDADF, HDDAF, and HDDAA motifs). In Y. pestis, alanine substitution in any of the four amino acids corresponding to GGDE rendered the diguanylate cyclase HmsT inactive (34). In S. enterica serovar Typhimurium, alanine substitutions in both of the glycine residues inactivated the diguanylate cyclase activity of AdrA (62). All of the substitution mutations that we introduced into scrG in combination with the phosphodiesterase-inactivating deletion removing the EAL domain allowed the production of proteins that were still capable of inhibiting laf expression. Thus, none of these amino acids seem critical for the activity associated with the GGDEF domain in the absence of the EAL domain.

GGDEF substitution mutations were also introduced into the full-length protein to test the effects on the EAL-associated activity. Biochemical analysis of another GGDEF-EAL protein (Caulobacter crescentus CC3396), which also has a nonconventional GGDEF motif, has shown that amino acids within the GGDEF motif can exert critical allosteric influence upon the phosphodiesterase activity of the molecule (13). The two alanine substitution mutations (D224A and F226A) that were constructed in the HDDDF region (corresponding to the GGDEF signature) produced proteins with activity similar to that of wild-type ScrG in the ability to activate laf gene expression in liquid. One mutation (D225A) caused reduced activity, although it did not eliminate the activity. Thus, none of these residues in the modified GGDEF motif in ScrG seem to have an essential role for phosphodiesterase activity; however, the D225A allele may contribute to or influence the phosphodiesterase activity.

How ScrG_EAL acts to influence the cellular nucleotide pool is not clear. Although assigning diguanylate cyclase activity to the GGDEF domain of ScrG is the simplest hypothesis, ScrG_EAL may act in other ways to affect the nucleotide pool. It might, for example, negatively interfere with other components in the signaling pathway. Perhaps ScrG_EAL overproduction leads to higher intracellular c-di-GMP levels as a consequence of interaction with other c-di-GMP-signaling proteins, as has been observed to be the case for a HD-GYP phosphodiesterase that physically interacts with several GGDEF proteins in Xanthomonas axonopodis pv. citri (3). Alternatively, ScrG_EAL production may impinge upon the concentration or balance of other cellular nucleotides. Clearly, further experiments with purified ScrG and derivatives are required to unambiguously define the biochemical activity (or activities) of ScrG.

What is the in vivo role of ScrG? The scrG::Tn5 allele affected swarming and colony morphology when transferred to the chromosome, although not greatly. However, on plates with higher agar or lower NaCl than our standard swarming conditions, effects on swarming could most readily be detected. Lateral flagellar gene expression and swarming motility decrease with decreasing NaCl concentration (L. L. McCarter, unpublished data). It seems likely that motility is so rapid on standard swarm plates that small decreases in mobility cannot be distinguished, whereas under conditions more restrictive for motility, the mutant phenotypes become articulated—although it is also possible that these particular conditions may alter the relative importance of ScrG and/or the balance of c-di-GMP in the cell. Like the Tn5 mutation, deletions in the chromosomal copy of scrG also failed to cause a detectable swarming phenotype when the mutant was analyzed on standard swarming medium. In fact, both deletion mutants (scrG and scrG_EAL) showed little observable defect under the low-salt conditions that revealed a phenotype for the Tn5 insertion mutant. However, swarming can be further restricted by supplementation with iron (40). On low-salt, iron-rich medium, the three scrG alleles (scrG::Tn5, scrG, and scrG_EAL) behaved similarly, displaying decreased production of lateral flagella and defective swarming. Thus, removal of the entire gene caused the same phenotype as deletion of the EAL domain or the Tn5 insertion in the EAL domain; therefore, ScrG probably acts as a phosphodiesterase in vivo under these conditions.

The Tn5 insertion, which occurs within the codon for aa 460 in the C-terminal EAL domain, caused a stronger mutant phenotype than did the scrG or scrG_EAL alleles. The phenotype of the scrG::Tn5 mutant (decreased swarming and increased colony roughness) mirrors the effects produced upon overexpression of the scrG_EAL construct. Both of these mutations eliminate the EAL domain to different extents. The scrG_EAL allele, which removes the coding sequence for residues 309 to 556, may produce a truncated protein with similar activity that is less stable or less active than the Tn5-induced truncation (which terminates at aa 460).

The potential degree of complexity, cross-talk, and specificity within the cellular network of GGDEF-EAL molecules is only now being fully appreciated. For example, the hierarchical involvement of numerous GGDEF domain proteins has been explored in S. enterica serovar Typhimurium and P. aeruginosa (23, 32, 35). It seems that there can be temporal and environmental control of GGDEF-EAL gene expression as well as enzymatic activity (reviewed in reference 53); furthermore, c-di-GMP may also act locally, the consequence of specific protein localization and sequestration of c-di-GMP with its target (46, 56). In summary, we observe that the loss of multiple GGDEF-EAL proteins in V. parahaemolyticus has cumulative consequences on swarming and colony morphology. Such observations support the idea that GGDEF-EAL proteins can constitute a network of signal-transducing proteins influencing the level of a small signaling molecule. Their net effect on the cellular concentration of c-di-GMP may serve to integrate diverse environmental and cellular signals to coordinate the production of cell surface molecules and accordingly to mediate cell interactions.

	ACKNOWLEDGMENTS

 
This work was supported by National Science Foundation grant MCB0315617.

	FOOTNOTES

 
* Corresponding author. Mailing address: Microbiology Department, The University of Iowa, Iowa City, IA 52242. Phone: (319) 335-9721. Fax: (319) 335-9721. E-mail: linda-mccarter{at}uiowa.edu

Published ahead of print on 30 March 2007.

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Aldridge, P., R. Paul, P. Goymer, P. Rainey, and U. Jenal. 2003. Role of the GGDEF regulator PleD in polar development of Caulobacter crescentus. Mol. Microbiol. 47:1695-1708.[CrossRef][Medline]
2. Amikam, D., and M. Benziman. 1989. Cyclic diguanylic acid and cellulose synthesis in Agrobacterium tumefaciens. J. Bacteriol. 171:6649-6655.[Abstract/Free Full Text]
3. Andrade, M. O., M. C. Alegria, C. R. Guzzo, C. Docena, M. C. Rosa, C. H. Ramos, and C. S. Farah. 2006. The HD-GYP domain of RpfG mediates a direct linkage between the Rpf quorum-sensing pathway and a subset of diguanylate cyclase proteins in the phytopathogen Xanthomonas axonopodis pv. citri. Mol. Microbiol. 62:537-551.[CrossRef][Medline]
4. Ausmees, N., H. Jonsson, S. Hoglund, H. Ljunggren, and M. Lindberg. 1999. Structural and putative regulatory genes involved in cellulose synthesis in Rhizobium leguminosarum bv. trifolii. Microbiology 145:1253-1262.[Abstract]
5. Beyhan, S., A. D. Tischler, A. Camilli, and F. H. Yildiz. 2006. Transcriptome and phenotypic responses of Vibrio cholerae to increased cyclic di-GMP level. J. Bacteriol. 188:3600-3613.[Abstract/Free Full Text]
6. Bobrov, A. G., O. Kirillina, and R. D. Perry. 2005. The phosphodiesterase activity of the HmsP EAL domain is required for negative regulation of biofilm formation in Yersinia pestis. FEMS Microbiol. Lett. 247:123-130.[CrossRef][Medline]
7. Bochner, B. R., and B. N. Ames. 1982. Complete analysis of cellular nucleotides by two-dimensional thin layer chromatography. J. Biol. Chem. 257:9759-9769.[Abstract/Free Full Text]
8. Boles, B. R., and L. L. McCarter. 2000. Insertional inactivation of genes encoding components of the sodium-type flagellar motor and switch of Vibrio parahaemolyticus. J. Bacteriol. 182:1035-1045.[Abstract/Free Full Text]
9. Boles, B. R., and L. L. McCarter. 2002. Vibrio parahaemolyticus scrABC, a novel operon affecting swarming and capsular polysaccharide regulation. J. Bacteriol. 184:5946-5954.[Abstract/Free Full Text]
10. Bomchil, N., P. Watnick, and R. Kolter. 2003. Identification and characterization of a Vibrio cholerae gene, mbaA, involved in maintenance of biofilm architecture. J. Bacteriol. 185:1384-1390.[Abstract/Free Full Text]
11. Chang, A. L., J. R. Tuckerman, G. Gonzalez, R. Mayer, H. Weinhouse, G. Volman, D. Amikam, M. Benziman, and M. A. Gilles-Gonzalez. 2001. Phosphodiesterase A1, a regulator of cellulose synthesis in Acetobacter xylinum, is a heme-based sensor. Biochemistry 40:3420-3426.[CrossRef][Medline]
12. Chow, W. Y., and D. E. Berg. 1988. Tn5tac1, a derivative of transposon Tn5 that generates conditional mutations. Proc. Natl. Acad. Sci. USA 85:6468-6472.[Abstract/Free Full Text]
13. Christen, M., B. Christen, M. Folcher, A. Schauerte, and U. Jenal. 2005. Identification and characterization of a cyclic di-GMP specific phosphodiesterase and its allosteric control by GTP. J. Biol. Chem. 281:32015-32024.[CrossRef]
14. Da Re, S., and J. M. Ghigo. 2006. A CsgD-independent pathway for cellulose production and biofilm formation in Escherichia coli. J. Bacteriol. 188:3073-3087.[Abstract/Free Full Text]
15. D'Argenio, D. A., M. W. Calfee, P. B. Rainey, and E. C. Pesci. 2002. Autolysis and autoaggregation in Pseudomonas aeruginosa colony morphology mutants. J. Bacteriol. 184:6481-6489.[Abstract/Free Full Text]
16. D'Argenio, D. A., and S. I. Miller. 2004. Cyclic di-GMP as a bacterial second messenger. Microbiology 150:2497-2502.[Abstract/Free Full Text]
17. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.[Abstract/Free Full Text]
18. Enos-Berlage, J., Z. T. Guvener, C. Keenan, and L. McCarter. 2005. Genetic determinants of biofilm formation in opaque and translucent Vibrio parahaemolyticus. Mol. Microbiol. 55:1160-1182.[CrossRef][Medline]
19. Enos-Berlage, J. L., and L. L. McCarter. 2000. Relation of capsular polysaccharide production and colonial cell organization to colony morphology in Vibrio parahaemolyticus. J. Bacteriol. 182:5513-5520.[Abstract/Free Full Text]
20. Galperin, M. Y. 2004. Bacterial signal transduction network in a genomic perspective. Environ. Microbiol. 6:552-567.[CrossRef][Medline]
21. Galperin, M. Y. 2006. Structural classification of bacterial response regulators: diversity of output domains and domain combinations. J. Bacteriol. 188:4169-4182.[Abstract/Free Full Text]
22. Galperin, M. Y., A. N. Nikolskaya, and E. V. Koonin. 2001. Novel domains of the prokaryotic two-component signal transduction systems. FEMS Microbiol. Lett. 203:11-21.[CrossRef][Medline]
23. Garcia, B., C. Latasa, C. Solano, F. Garcia-del Portillo, C. Gamazo, and I. Lasa. 2004. Role of the GGDEF protein family in Salmonella cellulose biosynthesis and biofilm formation. Mol. Microbiol. 54:264-277.[CrossRef][Medline]
24. Gjermansen, M., P. Ragas, C. Sternberg, S. Molin, and T. Tolker-Nielsen. 2005. Characterization of starvation-induced dispersion in Pseudomonas putida biofilms. Environ. Microbiol. 7:894-906.[CrossRef][Medline]
25. Gronewold, T. M., and D. Kaiser. 2002. act operon control of developmental gene expression in Myxococcus xanthus. J. Bacteriol. 184:1172-1179.[Abstract/Free Full Text]
26. Guvener, Z. T., and L. L. McCarter. 2003. Multiple regulators control capsular polysaccharide production in Vibrio parahaemolyticus. J. Bacteriol. 185:5431-5441.[Abstract/Free Full Text]
27. Hecht, G. B., and A. Newton. 1995. Identification of a novel response regulator required for the swarmer-to-stalked-cell transition in Caulobacter crescentus. J. Bacteriol. 177:6223-6229.[Abstract/Free Full Text]
28. Hickman, J. W., D. F. Tifrea, and C. S. Harwood. 2005. A chemosensory system that regulates biofilm formation through modulation of cyclic diguanylate levels. Proc. Natl. Acad. Sci. USA 102:14422-14427.[Abstract/Free Full Text]
29. Hisert, K. B., M. MacCoss, M. U. Shiloh, K. H. Darwin, S. Singh, R. A. Jones, S. Ehrt, Z. Zhang, B. L. Gaffney, S. Gandotra, D. W. Holden, D. Murray, and C. Nathan. 2005. A glutamate-alanine-leucine (EAL) domain protein of Salmonella controls bacterial survival in mice, antioxidant defence and killing of macrophages: role of cyclic diGMP. Mol. Microbiol. 56:1234-1245.[CrossRef][Medline]
30. Huang, B., C. B. Whitchurch, and J. S. Mattick. 2003. FimX, a multidomain protein connecting environmental signals to twitching motility in Pseudomonas aeruginosa. J. Bacteriol. 185:7068-7076.[Abstract/Free Full Text]
31. Jenal, U. 2004. Cyclic di-guanosine-monophosphate comes of age: a novel secondary messenger involved in