Allosteric Control of Cyclic di-GMP Signaling
2006; Elsevier BV; Volume: 281; Issue: 42 Linguagem: Inglês
10.1016/s0021-9258(19)84115-7
ISSN1083-351X
AutoresBeat Christen, Matthias Christen, Ralf Paul, Franziska F.-F. Schmid, Marc Folcher, Paul Jenoe, Markus Meuwly, Urs Jenal,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoCyclic di-guanosine monophosphate is a bacterial second messenger that has been implicated in biofilm formation, antibiotic resistance, and persistence of pathogenic bacteria in their animal host. Although the enzymes responsible for the regulation of cellular levels of c-di-GMP, diguanylate cyclases (DGC) and phosphodiesterases, have been identified recently, little information is available on the molecular mechanisms involved in controlling the activity of these key enzymes or on the specific interactions of c-di-GMP with effector proteins. By using a combination of genetic, biochemical, and modeling techniques we demonstrate that an allosteric binding site for c-di-GMP (I-site) is responsible for non-competitive product inhibition of DGCs. The I-site was mapped in both multi- and single domain DGC proteins and is fully contained within the GGDEF domain itself. In vivo selection experiments and kinetic analysis of the evolved I-site mutants led to the definition of an RXXD motif as the core c-di-GMP binding site. Based on these results and based on the observation that the I-site is conserved in a majority of known and potential DGC proteins, we propose that product inhibition of DGCs is of fundamental importance for c-di-GMP signaling and cellular homeostasis. The definition of the I-site binding pocket provides an entry point into unraveling the molecular mechanisms of ligand-protein interactions involved in c-di-GMP signaling and makes DGCs a valuable target for drug design to develop new strategies against biofilm-related diseases. Cyclic di-guanosine monophosphate is a bacterial second messenger that has been implicated in biofilm formation, antibiotic resistance, and persistence of pathogenic bacteria in their animal host. Although the enzymes responsible for the regulation of cellular levels of c-di-GMP, diguanylate cyclases (DGC) and phosphodiesterases, have been identified recently, little information is available on the molecular mechanisms involved in controlling the activity of these key enzymes or on the specific interactions of c-di-GMP with effector proteins. By using a combination of genetic, biochemical, and modeling techniques we demonstrate that an allosteric binding site for c-di-GMP (I-site) is responsible for non-competitive product inhibition of DGCs. The I-site was mapped in both multi- and single domain DGC proteins and is fully contained within the GGDEF domain itself. In vivo selection experiments and kinetic analysis of the evolved I-site mutants led to the definition of an RXXD motif as the core c-di-GMP binding site. Based on these results and based on the observation that the I-site is conserved in a majority of known and potential DGC proteins, we propose that product inhibition of DGCs is of fundamental importance for c-di-GMP signaling and cellular homeostasis. The definition of the I-site binding pocket provides an entry point into unraveling the molecular mechanisms of ligand-protein interactions involved in c-di-GMP signaling and makes DGCs a valuable target for drug design to develop new strategies against biofilm-related diseases. A global signaling network that relies on the production of the second messenger cyclic diguanylic acid has recently been discovered in bacteria (1Jenal U. Curr. Opin. Microbiol. 2004; 7: 185-191Crossref PubMed Scopus (177) Google Scholar, 2Romling U. Gomelsky M. Galperin M.Y. Mol. Microbiol. 2005; 57: 629-639Crossref PubMed Scopus (549) Google Scholar). The c-di-GMP 3The abbreviations used are: c-di-GMP, cyclic diguanylic acid; pGpG, linear diguanylic acid; LB, Luria broth; DGC, diguanylate cyclase; PDE, phosphodiesterase; H6, hexa-histidine tag; rdar, red, dry, and rough; IPTG, isopropyl 1-thio-β-d-galactopyranoside; DgcA, diguanylate cyclase A; PdeA, phosphodiesterase A; CR, Congo Red; AC, adenylate cyclase; GC, guanylate cyclase. 3The abbreviations used are: c-di-GMP, cyclic diguanylic acid; pGpG, linear diguanylic acid; LB, Luria broth; DGC, diguanylate cyclase; PDE, phosphodiesterase; H6, hexa-histidine tag; rdar, red, dry, and rough; IPTG, isopropyl 1-thio-β-d-galactopyranoside; DgcA, diguanylate cyclase A; PdeA, phosphodiesterase A; CR, Congo Red; AC, adenylate cyclase; GC, guanylate cyclase. system emerges as a regulatory mastermind orchestrating multicellular behavior and biofilm formation in a wide variety of bacteria (2Romling U. Gomelsky M. Galperin M.Y. Mol. Microbiol. 2005; 57: 629-639Crossref PubMed Scopus (549) Google Scholar). In addition, c-di-GMP signaling also plays a role in bacterial virulence and persistence (3Brouillette E. Hyodo M. Hayakawa Y. Karaolis D.K.R. Malouin F. Antimicrob. Agents Chemother. 2005; 49: 3109-3113Crossref PubMed Scopus (75) Google Scholar, 4Hisert K.B. MacCoss M. Shiloh M.U. Darwin K.H. Singh S. Jones R.A. Ehrt S. Zhang Z.Y. Gaffney B.L. Gandotra S. Holden D.W. Murray D. Nathan C. Mol. Microbiol. 2005; 56: 1234-1245Crossref PubMed Scopus (118) Google Scholar, 5Kulesekara H. Lee V. Brencic A. Liberati N. Urbach J. Miyata S. Lee D.G. Neely A.N. Hyodo M. Hayakawa Y. Ausubel F.M. Lory S. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 2839-2844Crossref PubMed Scopus (436) Google Scholar, 6Lestrate P. Dricot A. Delrue R.M. Lambert C. Martinelli V. De Bolle X. Letesson J.J. Tibor A. Infect. Immun. 2003; 71: 7053-7060Crossref PubMed Scopus (80) Google Scholar, 7Tischler A.D. Lee S.H. Camilli A. J. Bacteriol. 2002; 184: 4104-4113Crossref PubMed Scopus (49) Google Scholar). The broad importance of this novel signaling molecule in pathogenic and non-pathogenic bacteria calls for careful analysis of the molecular mechanisms that control cellular levels of c-di-GMP and regulate its downstream targets. c-di-GMP is formed by the condensation of two GTP molecules (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar, 9Ross P. Weinhouse H. Aloni Y. Michaeli D. Weinbergerohana P. Mayer R. Braun S. Devroom E. Vandermarel G.A. Vanboom J.H. Benziman M. Nature. 1987; 325: 279-281Crossref PubMed Scopus (801) Google Scholar, 10Ryjenkov D.A. Tarutina M. Moskvin O.V. Gomelsky M. J. Bacteriol. 2005; 187: 1792-1798Crossref PubMed Scopus (461) Google Scholar) and is hydrolyzed to GMP via the linear intermediate pGpG (11Ross P. Mayer R. Benziman M. Microbiol. Rev. 1991; 55: 35-58Crossref PubMed Google Scholar, 12Schmidt A.J. Ryjenkov D.A. Gomelsky M. J. Bacteriol. 2005; 187: 4774-4781Crossref PubMed Scopus (446) Google Scholar, 13Tamayo R. Tischler A.D. Camilli A. J. Biol. Chem. 2005; 280: 33324-33330Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar, 14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar). Two widespread and highly conserved bacterial protein domains have been implicated in the synthesis and hydrolysis of c-di-GMP, respectively (15Tal R. Wong H.C. Calhoon R. Gelfand D. Fear A.L. Volman G. Mayer R. Ross P. Amikam D. Weinhouse H. Cohen A. Sapir S. Ohana P. Benziman M. J. Bacteriol. 1998; 180: 4416-4425Crossref PubMed Google Scholar). The breakdown of c-di-GMP is catalyzed by the EAL domain (12Schmidt A.J. Ryjenkov D.A. Gomelsky M. J. Bacteriol. 2005; 187: 4774-4781Crossref PubMed Scopus (446) Google Scholar, 13Tamayo R. Tischler A.D. Camilli A. J. Biol. Chem. 2005; 280: 33324-33330Abstract Full Text Full Text PDF PubMed Scopus (218) Google Scholar, 14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar), and the diguanylate cyclase (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar) activity resides in the GGDEF domain (10Ryjenkov D.A. Tarutina M. Moskvin O.V. Gomelsky M. J. Bacteriol. 2005; 187: 1792-1798Crossref PubMed Scopus (461) Google Scholar, 16Paul R. Weiser S. Amiot N.C. Chan C. Schirmer T. Giese B. Jenal U. Genes Dev. 2004; 18: 715-727Crossref PubMed Scopus (487) Google Scholar). The highly conserved amino acid sequence GG(D/E)EF forms part of the catalytically active site (A-site) of the DGC enzyme (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar). In agreement with this, mutations that change the GG(D/E)EF motif generally abolish the activity of the respective proteins (14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar, 16Paul R. Weiser S. Amiot N.C. Chan C. Schirmer T. Giese B. Jenal U. Genes Dev. 2004; 18: 715-727Crossref PubMed Scopus (487) Google Scholar, 17Aldridge P. Paul R. Goymer P. Rainey P. Jenal U. Mol. Microbiol. 2003; 47: 1695-1708Crossref PubMed Scopus (190) Google Scholar, 18Kirillina O. Fetherston J.D. Bobrov A.G. Abney J. Perry R.D. Mol. Microbiol. 2004; 54: 75-88Crossref PubMed Scopus (193) Google Scholar). GGDEF domains are often found associated with sensor domains, arguing that DGC activity is controlled by direct signal input through these domains (1Jenal U. Curr. Opin. Microbiol. 2004; 7: 185-191Crossref PubMed Scopus (177) Google Scholar). The best understood example for controlled activation of a DGC is the response regulator PleD, which constitutes a timing device for Caulobacter crescentus pole development (17Aldridge P. Paul R. Goymer P. Rainey P. Jenal U. Mol. Microbiol. 2003; 47: 1695-1708Crossref PubMed Scopus (190) Google Scholar, 19Aldridge P. Jenal U. Mol. Microbiol. 1999; 32: 379-391Crossref PubMed Scopus (102) Google Scholar, 20Hecht G.B. Newton A. J. Bacteriol. 1995; 177: 6223-6229Crossref PubMed Google Scholar). PleD is activated during C. crescentus development by phosphorylation of an N-terminal receiver domain and, as a result, sequesters to the differentiating cell pole (17Aldridge P. Paul R. Goymer P. Rainey P. Jenal U. Mol. Microbiol. 2003; 47: 1695-1708Crossref PubMed Scopus (190) Google Scholar, 19Aldridge P. Jenal U. Mol. Microbiol. 1999; 32: 379-391Crossref PubMed Scopus (102) Google Scholar). An additional layer of control was suggested by the crystal structure of PleD solved recently in complex with c-di-GMP (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar) (Fig. 1). A c-di-GMP binding site was identified in the crystal, spatially separated from the catalytically active site (A-site). Two mutually intercalating c-di-GMP molecules were found tightly bound to this site, at the interface between the GGDEF and the central receiver-like domain of PleD (Fig. 1). Based on the observation that PleD activity shows a strong non-competitive product inhibition, it was proposed that this site might constitute an allosteric binding site (I-site) (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar). Based on the observation that functionally important residues of the PleD I-site are highly conserved in a majority of GGDEF proteins listed in the data base, we tested the hypothesis that allosteric product inhibition is a general regulatory principle of bacterial diguanylate cyclases. Strains, Plasmids, and Media—Escherichia coli and Salmonella enterica serovar Typhimurium strains were grown in Luria broth (LB). C. crescentus strains were grown in complex peptone yeast extract (21Ely B. Methods Enzymol. 1991; 204: 372-384Crossref PubMed Scopus (360) Google Scholar). For DGC activity assays in vivo, E. coli was plated onto LB Congo Red plates (Sigma, 50 μg/ml). To determine the IPTG induction phenotype, 3 μl of a liquid log phase culture was spotted onto LB Congo Red plates without and with 1 mm IPTG. Biofilm formation was quantified after overnight growth by staining with 1% Crystal Violet as described (22O'Toole G.A. Pratt L.A. Watnick P.I. Newman D.K. Weaver V.B. Kolter R. Methods Enzymol. 1999; 34: 586-595Google Scholar). Motility phenotypes were determined using LB or peptone yeast extract motility plates containing 0.3% Difco-Agar. The exact procedure of strain and plasmid construction is available on request. Random I-site Tetrapeptide Library—The dgcA gene (CC3285) was amplified by PCR using primers #1006 and #1007 (for primer list see supplemental text). The PCR product was digested with NdeI and XhoI and cloned into pET21a (Novagen). In a next step a dgcAΔRESD allele with a silent PstI restriction site was generated by splicing with overlapping extension PCR using primers #1129, #670, and #1132. The resulting PCR product was digested with NdeI and XhoI and cloned into pET42b (Novagen) to produce pET42::dgcAΔRESD. The PstI/XhoI fragment of pET42b::dgcAΔRESD was replaced by 20 independent PCR products, which had been generated using pET42b::dgcAΔRESD as a template and primers #1131 and #670. The resulting 20 independent random libraries were individually transformed into E. coli BL21 and screened on Congo Red plates (LB plates supplemented with 50 μg/ml Congo Red). As a control reaction, the deleted I-site was reverted back to the wild-type RESD motif by cloning the PCR product generated with primers #1130 and #670 into the PstI and XhoI site of pET42b::dgcAΔRESD. Diguanylate Cyclase and Phosphodiesterase Activity Assays— DGC reactions were performed at 30 °C with 0.5 μm purified hexahistidine-tagged DgcA or 5 μm PleD in DGC reaction buffer containing 250 mm NaCl, 25 mm Tris-Cl, pH 8.0, 5 mm β-mercaptoethanol, and 20 mm MgCl2. For inhibition assays the protein was preincubated with different concentrations of c-di-GMP (1-100 μm) for 2 min at 30 °C before 100 μm [33P]GTP (Amersham Biosciences) was added. The reaction was stopped at regular time intervals by adding an equal volume of 0.5 m EDTA, pH 8.0. DGC/PDE tandem assays were carried out using 1 μm hexahistidine-tagged DgcA, which was preincubated for 2 min in the presence or absence of 4.5 μm hexahistidine-tagged phosphodiesterase PdeA. The reaction was started by adding 100 μm [33P]GTP. The reactions were stopped at regular time intervals of 15 s by adding equal volumes of 0.5 m EDTA, pH 8.0, and their nucleotide composition was analyzed as described below. Initial velocity (Vo) and inhibition constants were determined by plotting the corresponding nucleotide concentration versus time and by fitting the curve according to allosteric product inhibited Michaelis-Menten kinetics with the program ProFit 5.6.7 (with fit function [c-di-GMP]t = a(1)*t/(a(2) + t), where the initial velocity Vo is defined as a(1)/a(2)) using the Levenberg-Marquardt algorithm. Ki values were determined by plotting Vo versus c-di-GMP concentration and using the following fit function, Vo[c-di-GMP] = Vo[c-di-GMP] = 0 *(1 - ([c-di-GMP]/(Ki + [c-di-GMP])). Polyethyleneimine Cellulose Chromatography—Samples were dissolved in 5 μl of running buffer containing 1:1.5 (v/v) saturated NH4SO4 and 1.5 m KH2PO4, pH 3.60, and blotted on Polygram® CEL 300 polyethyleneimine cellulose TLC plates (Macherey-Nagel). Plates were developed in 1:1.5 (v/v) saturated NH4SO4 and 1.5 m KH2PO4, pH 3.60 (Rf(c-di-GMP) 0.2, Rf(pGpG) 0.4), dried, and exposed on a storage phosphor imaging screen (Amersham Biosciences). The intensity of the various radioactive species was calculated by quantifying the intensities of the relevant spots using ImageJ software version 1.33. Vo and Ki were determined with the Software ProFit 5.6.7. UV Cross-linking with [33P]c-di-GMP—The 33P-labeled c-di-GMP was produced enzymatically using [33P]GTP (3000 Ci/mmol) and purified according to a previous study (14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar). Protein samples were incubated for 10 min on ice in DGC reaction buffer (25 mm Tris-HCl, pH 8.0, 250 mm NaCl, 10 mm MgCl2,5 mm β-mercaptoethanol) together with 1 μm c-di-GMP and 33P-radiolabeled c-di-GMP (0.75 μCi, 6000 Ci/mmol). Samples were then irradiated at 254 nm for 20 min in an ice-cooled, parafilm-wrapped 96-well aluminum block in an RPR-100 photochemical reactor with a UV lamp RPR-3500 (Southern New England Ultraviolet Co.). After irradiation, samples were mixed with 2× SDS-PAGE sample buffer (250 mm Tris-HCl at pH 6.8, 40% glycerol, 8% SDS, 2.4 m β-mercaptoethanol, 0.06% bromphenol blue, 40 mm EDTA) and heated for 5 min at 95 °C. Labeled proteins were separated by SDS-PAGE and quantified by autoradiography. Nucleotide Extraction and Analysis—2.0 ml of E. coli cell cultures (A600 0.4) were harvested by centrifugation, and supernatant was discarded. The cell pellet was dissolved in 200 μlof0.5 m formic acid, and nucleotides were extracted for 10 min at 4 °C. Insoluble cell components were then pelleted, and the supernatant was directly analyzed by chromatography. Nucleotides were extracted and separated according to a previous study (23Ochi Y. Hosoda S. Hachiya T. Yoshimura M. Miyazaki T. Kajita Y. Acta Endocrinol. 1981; 98: 62-67Crossref PubMed Scopus (6) Google Scholar) on a 125/4 Nucleosil 4000-1 polyethyleneimine column (Macherey-Nagel) using the SMART-System (Amersham Biosciences). The nucleotide peak corresponding to c-di-GMP was verified by co-elution with a chemically synthesized c-di-GMP standard. DgcA Protein Expression Levels—DgcA protein expression levels in E. coli BL21 were determined by Western blot analysis using Anti-His(C-Term) antibody (Invitrogen) and horseradish peroxidase conjugate of goat anti-mouse IgG (Invitrogen) as secondary antibody. The protein concentration was determined by measuring the intensities of the relevant spots using ImageJ software version 1.33. Signals were calibrated to defined concentrations of purified wild-type DgcA. Molecular Modeling of PleD—All-atom simulations were carried out using the CHARMM (24Brooks B.R. Bruccoleri R.E. Olafson B.D. States D.J. Swaminathan S. Karplus M. J. Comput. Chem. 1983; 4: 187-217Crossref Scopus (13939) Google Scholar) program and the CHARMM22/27 force field (25MacKerell A.D. Bashford D. Bellott M. Dunbrack R.L. Evanseck J.D. Field M.J. Fischer S. Gao J. Guo H. Ha S. Joseph-McCarthy D. Kuchnir L. Kuczera K. Lau F.T.K. Mattos C. Michnick S. Ngo T. Nguyen D.T. Prodhom B. Reiher W.E. Roux B. Schlenkrich M. Smith J.C. Stote R. Straub J. Watanabe M. Wiorkiewicz-Kuczera J. Yin D. Karplus M. J. Phys. Chem. B. 1998; 102: 3586-3616Crossref PubMed Scopus (11685) Google Scholar). For additional information see the supplemental material. Feedback Inhibition of the PleD Diguanylate Cyclase Requires Binding of c-di-GMP to the I-site—The PleD crystal structure indicated the existence of an allosteric binding pocket (I-site) at the interface of the GGDEF and REC2 domains (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar). Binding of a c-di-GMP dimer in the I-site is mediated by specific electrostatic interactions with charged residues of the GGDEF and REC2 domain (Fig. 1). To provide evidence for c-di-GMP binding to the I-site pocket in solution, trypsin digests were performed with purified PleD protein (5 μm) in the presence or absence of c-di-GMP (25 μm). The resulting peptide fragments were separated on a C18 column and analyzed by matrix-assisted laser desorption ionization time-of-flight. Both chromatograms were identical, with the exception of two peaks that were only detected in the absence of ligand but were protected when c-di-GMP present during tryptic digest (supplemental Fig. S1). One of the two peptides (T47, retention time 25.6 min) was identified by mass spectrometry and corresponds to the amino acids 354-359 (supplemental Fig. S1), arguing that c-di-GMP specifically protects from trypsin cleavage at Arg-359. To provide additional evidence for ligand binding in solution, we performed UV cross-linking assays using 33P-labeled c-di-GMP (14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar). Residues Arg-148 and Arg-178 of the REC2 domain and Arg-359, Asp-362, and Arg-390 of the GGDEF domain were replaced with alanine, and the resulting protein variants were analyzed. As shown in Fig. 2, mutating I-site residues of the GGDEF domain abolished (ΔR359ΔD362) or strongly reduced (R359A and R390A) c-di-GMP binding. In contrast, mutations in the A-site (E370Q, E371Q, and EE370GG), which completely abolished enzymatic activity (Table 1), had no effect on c-di-GMP binding (Fig. 2), indicating that labeling with radioactive c-di-GMP results from ligand binding at the I-site. Although mutations R359A, R359V, ΔR359ΔD362, and D362A all showed a dramatically reduced or complete loss of enzymatic activity, mutant R390A showed wild-type-like DGC activity (Table 1). In agreement with the reduced binding of c-di-GMP (Fig. 2), the Ki of mutant R390A was increased ∼20-fold (Table 1). PleD proteins harboring mutations in the REC2 portion of the I-site (R148A and R178A) showed an increased binding of c-di-GMP (Fig. 2) and slightly lower Ki values than wild type (Table 1). Surprisingly, R148A/R178A single and double mutants displayed a 5- to 20-fold higher DGC activity compared with wild-type PleD (Table 1). Finally, c-di-GMP binding was normal in mutant proteins that either lacked the REC1 receiver domain or had a bulky tryptophan residue introduced at the REC2-GGDEF interface (G194W, Fig. 2). Together these results implied that the structural requirements for c-di-GMP binding are contained within the GGDEF domain of PleD and that residues Arg-359, Asp-362, and Arg-390 form the core of an allosteric binding pocket for c-di-GMP.TABLE 1Kinetic analysis of PleD mutantsProteinVoΔVoKiΔKiμmol c-di-GMP/(μmol protein*min)μmPleD wild type0.202± 0.0235.8± 1.0PleDR359A0.005NDaND, not determined>100NDPleDR359V0.0NDPleDΔ359Δ3620.0NDPleDD362A0.0NDPleDR390A0.076± 0.007115.0± 18.1PleDR148A0.822± 0.0202.8± 1.2PleDR178A0.918± 0.2923.6± 0.1PleDR148AR178A3.75± 0.432.9± 0.6PleDG194W0.161± 0.0056.3± 1.9PleDEE370GG0.0NDPleDE370Q0.0NDPleDE371Q0.0NDa ND, not determined Open table in a new tab Evidence for an in Vivo Role of I-site-mediated Feedback Control—To test a possible role for feedback inhibition of diguanylate cyclases in vivo, we developed a simple assay based on the observation that in E. coli and other Enterobacteriaceae increased cellular levels of c-di-GMP correlate with Congo Red (CR) staining of colonies on plates (28Garcia B. Latasa C. Solano C. Portillo F.G. Gamazo C. Lasa I. Mol. Microbiol. 2004; 54: 264-277Crossref PubMed Scopus (190) Google Scholar). Low level expression (in the absence of the inducer IPTG) of active pleD alleles caused a red colony phenotype in the E. coli B strain BL21, whereas cells expressing inactive pleD alleles under the same conditions stained white (Fig. 3). Interestingly, PleD mutants with dramatically different diguanylate cyclase activities in vitro showed only minor differences of CR staining in vivo. For instance, PleDR148A/R178A, which showed a 20-fold increased activity (Table 1), or PleD*, a constitutively active mutant of PleD several 100-fold more active than wild-type (9Ross P. Weinhouse H. Aloni Y. Michaeli D. Weinbergerohana P. Mayer R. Braun S. Devroom E. Vandermarel G.A. Vanboom J.H. Benziman M. Nature. 1987; 325: 279-281Crossref PubMed Scopus (801) Google Scholar), caused virtually identical CR values like PleD wild type (Fig. 3). In contrast, expression of the feedback inhibition mutant PleDR390A resulted in a significantly higher CR staining even though its in vitro DGC activity was lower than wild-type PleD (Table 1). This argued that in vivo steady-state concentrations of c-di-GMP were determined mainly by the PleD inhibition constant (as opposed to the overall activity of the enzyme) and that a functional I-site is critical for DGC control in vivo. DgcA, a Single Domain Diguanylate Cyclase, Is Subject to Allosteric Product Inhibition—Sequence alignments of >1000 annotated GGDEF domain proteins revealed that that I-site residues Arg-359 and Asp-362 of PleD are highly conserved. 57% of the proteins containing a GGDEF domain and 27% of GGDEF/EAL composite proteins possess this motif. This suggested that c-di-GMP product inhibition could be a general regulatory mechanism of bacterial diguanylate cyclases. To test this, hexahistidine-tagged derivatives of two C. crescentus GGDEF domain proteins were analyzed biochemically with respect to their DGC activities and c-di-GMP binding properties. Purified DgcA (diguanylate cyclase A, CC3285), a soluble, single domain GGDEF protein that lacks an obvious N-terminal input domain, showed strong diguanylate cyclase activity (Fig. 5A). DgcA has an RESD motive five amino acids upstream of the conserved GGDEF active site and was readily labeled with [33P]c-di-GMP in a cross-linking experiment (Fig. 4). Consistent with this, DgcA showed strong feedback inhibition (Fig. 5A) with its Ki (1 μm) being in the same range as the inhibition constant determined for PleD (8Chan C. Paul R. Samoray D. Amiot N.C. Giese B. Jenal U. Schirmer T. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 17084-17089Crossref PubMed Scopus (362) Google Scholar). In contrast, the GGDEF domain of PdeA (phosphodiesterase A, CC3396), which lacks catalytic activity (14Christen M. Christen B. Folcher M. Schauerte A. Jenal U. J. Biol. Chem. 2005; 280: 30829-30837Abstract Full Text Full Text PDF PubMed Scopus (372) Google Scholar), had no conserved I-site residues and did not bind radiolabeled c-di-GMP (Fig. 4). Thus, specific binding of c-di-GMP correlated with the presence of a conserved I-site motif RXXD (Fig. 4).FIGURE 4UV cross-linking of different GGDEF domains with 33P-labeled c-di-GMP. A, Coomassie-stained SDS-PAGE and B, autoradiograph of BSA (control), PleDΔREC1, DgcA, and the isolated GGDEF domain of the c-di-GMP-specific phosphodiesterase PdeA (CC3396) after UV cross-linking with [33P]c-di-GMP. C, alignment of I- and A-site sequence of PleD, DgcA, and PdeA. I-site (RXXD) and A-site residues (GGDEF) are marked in black and gray, respectively.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Diguanylate cyclase activity assays revealed strong and rapid product inhibition of DgcA. DgcA alone was able to convert only a small fraction of the available GTP substrate pool into the product c-di-GMP (Vo = 2.8 μmol of c-di-GMP μmol protein-1 min-1) (Fig. 5B). In contrast, GTP consumption and conversion into c-di-GMP and pGpG was rapid (Vo = 43.0 μmol of c-di-GMP μmol protein-1 min-1) and almost complete when the PDE CC3396 was added in excess to the enzymatic reaction (Fig. 5B). This argued that c-di-GMP feedback inhibition is abolished in a sequential DGC-PDE reaction, because the steady-state concentration of the inhibitor c-di-GMP is kept low by continuous degradation of c-di-GMP into the linear dinucleotide pGpG. As a consequence of rapid feedback inhibition, the experimentally determined Vo values of the DGC reaction are generally underestimated. In conclusion, these results strengthen the view that allosteric product inhibition is a general principle of diguanylate cyclases and that high affinity binding of c-di-GMP requires an RXXD I-site motif positioned in close proximity to the active site. Development of an in Vivo Assay to Genetically Probe Allosteric Control of DgcA—DGCs from different bacterial species have been shown to be functionally interchangeable (17Aldridge P. Paul R. Goymer P. Rainey P. Jenal U. Mol. Microbiol. 2003; 47: 1695-1708Crossref PubMed Scopus (190) Google Scholar, 26Ausmees N. Mayer R. Weinhouse H. Volman G. Amikam D. Benziman M. Lindberg M. FEMS Microbiol. Lett. 2001; 204: 163-167Crossref PubMed Google Scholar, 27Simm R. Fetherston J.D. Kader A. Romling U. Perry R.D. J. Bacteriol. 2005; 187: 6816-6823Crossref PubMed Scopus (64) Google Scholar). To determine if DgcA is active in vivo we expressed a plasmid-based copy of the dgcA gene in C. crescentus, S. enterica, and Escherichia coli B and tested the respective strains for the phenotypes known to result from increased cellular levels of c-di-GMP (17Aldridge P. Paul R. Goymer P. Rainey P. Jenal U. Mol. Microbiol. 2003; 47: 1695-1708Crossref PubMed Scopus (190) Google Scholar, 26Ausmees N. Mayer R. Weinhouse H. Volman G. Amikam D. Benziman M. Lindberg M. FEMS Microbiol. Lett. 2001; 204: 163-167Crossref PubMed Google Scholar, 27Simm R. Fetherston J.D. Kader A. Romling U. Perry R.D. J. Bacteriol. 2005; 187: 6816-6823Crossref PubMed Scopus (64) Google Scholar). Consistent with these earlier findings, expression of dgcA strongly inhibited flagellar-based motility in all three organisms, dramatically increased the ability of S. enterica and E. coli for surface colonization, and produced the characteristic red, dry, and rough (rdar) colony morphotype when plated on CR plates (Fig. 6, A-F) (29Zogaj X. Nimtz M. Rohde M. Bokranz W. Romling U. Mol. Microbiol. 2001; 39: 1452-1463Crossref PubMed Scopus (707) Google Scholar). The red phenotype provided the basis for a visual genetic screen on CR plates. Under these conditions, cells producing active DgcA variants would produce dark red single colonies, whereas cells producing inactive DgcA mutants would remain white. This prompted us to use the CR scr
Referência(s)