Portal de Periódicos da CAPES

Bacillus subtilis DesR Functions as a Phosphorylation-activated Switch to Control Membrane Lipid Fluidity

2004; Elsevier BV; Volume: 279; Issue: 38 Linguagem: Inglês

10.1074/jbc.m405150200

1083-351X

Larisa E. Cybulski, Gloria del Solar, Patricio O. Craig, Manuel Espinosa, Diego de Mendoza,

Microbial Metabolic Engineering and Bioproduction

The Des pathway of Bacillus subtilis regulates the synthesis of the cold-shock induced membrane-bound enzyme Δ5-fatty acid desaturase (Δ5-Des). A central component of the Des pathway is the response regulator, DesR, which is activated by a membrane-associated kinase, DesK, in response to a decrease in membrane lipid fluidity. Despite genetic and biochemical studies, specific details of the interaction between DesR and the DNA remain unknown. In this study we show that only the phosphorylated form of protein DesR is able to bind to a regulatory region immediately upstream of the promoter of the Δ5-Des gene (Pdes). Phosphorylation of the regulatory domain of dimeric DesR promotes, in a cooperative fashion, the hierarchical occupation of two adjacent, non-identical, DesR-P DNA binding sites, so that there is a shift in the equilibrium toward the tetrameric active form of the response regulator. Subsequently, this phosphorylation signal propagation leads to the activation of the des gene through recruitment of RNA polymerase to Pdes. This is the first dissected example of a transcription factor functioning as a phosphorylation-activated switch for a cold-shock gene, allowing the cell to optimize the fluidity of membrane phospholipids. The Des pathway of Bacillus subtilis regulates the synthesis of the cold-shock induced membrane-bound enzyme Δ5-fatty acid desaturase (Δ5-Des). A central component of the Des pathway is the response regulator, DesR, which is activated by a membrane-associated kinase, DesK, in response to a decrease in membrane lipid fluidity. Despite genetic and biochemical studies, specific details of the interaction between DesR and the DNA remain unknown. In this study we show that only the phosphorylated form of protein DesR is able to bind to a regulatory region immediately upstream of the promoter of the Δ5-Des gene (Pdes). Phosphorylation of the regulatory domain of dimeric DesR promotes, in a cooperative fashion, the hierarchical occupation of two adjacent, non-identical, DesR-P DNA binding sites, so that there is a shift in the equilibrium toward the tetrameric active form of the response regulator. Subsequently, this phosphorylation signal propagation leads to the activation of the des gene through recruitment of RNA polymerase to Pdes. This is the first dissected example of a transcription factor functioning as a phosphorylation-activated switch for a cold-shock gene, allowing the cell to optimize the fluidity of membrane phospholipids. All organisms must communicate with their environment to survive. Bacterial cells monitor external conditions and process this information to give the most appropriate response. Two-component regulatory systems have emerged as a paradigm for adaptive responses. In its simplest form, a two-component system contains a sensor (histidine kinase) and a response regulator (often a transcriptional activator) (1Hoch J.A. Curr. Opin. Microbiol. 2000; 3: 165-170Crossref PubMed Scopus (632) Google Scholar). Changes in the environment result in phosphorylation of the sensor followed by transphosphorylation onto the response regulator (1Hoch J.A. Curr. Opin. Microbiol. 2000; 3: 165-170Crossref PubMed Scopus (632) Google Scholar, 2Fabret C. Feher V.A. Hoch J.A. J. Bacteriol. 1999; 181: 1975-1983Crossref PubMed Google Scholar). Although kinase-response regulator pairs of this type were frequently reported as governors of a wide variety of pathways in response to a myriad of signals (3Bourret R. Borkovich K. Simon M. Annu. Rev. Biochem. 1991; 60: 401-441Crossref PubMed Scopus (390) Google Scholar, 4Stock J. Surette M. Levit M. Park P. Hoch J. Silhavy T. Two Component Signal Transduction. American Society for Microbiology, Washington, D. C.1995: 25-51Google Scholar, 5Stock A. Robinson V. Goudreau P. Annu. Rev. Biochem. 2000; 69: 183-215Crossref PubMed Scopus (2431) Google Scholar), the requirement of this system to control gene expression during cold-shock has only recently been discovered in Bacillus subtilis (6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar) and cyanobacteria (7Suzuki I. Los D.A. Kanesaki Y. Mikami K. Murata N. EMBO J. 2000; 19: 1327-1334Crossref PubMed Scopus (219) Google Scholar). Cold shock is a stress condition that adversely affects the growth of poikilothermic organisms, such as bacteria and plants. Thus, understanding the mechanisms by which these organisms perceive low temperature signals and transmit this information to the cellular machinery to activate adaptive responses is of fundamental importance to biology. Bacteria and most (if not all) poikilothermic organisms have to remodel the membrane lipid composition to survive at low temperatures. B. subtilis (6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar) and Synechocystis (7Suzuki I. Los D.A. Kanesaki Y. Mikami K. Murata N. EMBO J. 2000; 19: 1327-1334Crossref PubMed Scopus (219) Google Scholar) respond to a decrease in the ambient growth temperature by introducing double bonds into the acyl chains of their membrane phospholipids by membrane-bound acyl lipid desaturases. This post-synthetic modification of the saturated acyl chains seems to be designed to ameliorate the effect of temperature changes on the physical state of membrane phospholipids. In Synechocystis it was found that inactivation of two histidine kinases moderates the low-temperature induction of the genes coding for Δ6 and ω-3 desaturase (7Suzuki I. Los D.A. Kanesaki Y. Mikami K. Murata N. EMBO J. 2000; 19: 1327-1334Crossref PubMed Scopus (219) Google Scholar). However, the transcriptional regulators of these genes were not yet identified. A better understood example of perception and transduction of low temperature signals is the Des pathway of B. subtilis, which regulates the expression of the acyl lipid desaturase, Δ5-Des. This pathway responds to a decrease in growth temperature by enhancing the expression of the des gene coding for Δ5-Des (8Aguilar P.S. Cronan Jr., J.E. de Mendoza D. J. Bacteriol. 1998; 180: 2194-2200Crossref PubMed Google Scholar, 9Aguilar P.S. Lopez P. de Mendoza D. J. Bacteriol. 1999; 181: 7028-7033Crossref PubMed Google Scholar, 10Díaz A. Mansilla M.C. Vila A. de Mendoza D. J. Biol. Chem. 2002; 277: 48099-48106Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar, 11Altabe S. Aguilar P. Caballero G. de Mendoza D. J. Bacteriol. 2003; 185: 3228-3231Crossref PubMed Scopus (66) Google Scholar). The Des pathway is uniquely and stringently controlled by the two-component system, DesK/DesR (6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar). DesK is a histidine kinase located in the membrane and DesR is a 22.18-kDa cytoplasmic response regulator that binds specifically to the promoter region it controls. Response regulators have been classified according to similarities in the DNA binding domain. DesR belongs to the NarL group, in which the C-terminal binding domain contains a helix-turn-helix motif (2Fabret C. Feher V.A. Hoch J.A. J. Bacteriol. 1999; 181: 1975-1983Crossref PubMed Google Scholar). Based on the two-component system paradigm and on previous work, we have proposed that the two proteins communicate via a series of phosphorylation and phosphotransfer reactions. DesK senses a decrease in membrane lipid fluidity and is autophosphorylated by intracellular ATP at His-188 (12Cybulski L. Albanesi D. Mansilla M.C. Altabe S. de Mendoza D. Mol. Microbiol. 2002; 45: 1379-1388Crossref PubMed Scopus (98) Google Scholar, 13Albanesi D. Mansilla M.C. de Mendoza D. J. Bacteriol. 2004; 186: 2655-2663Crossref PubMed Scopus (84) Google Scholar). We have recently demonstrated that DesK-P is able to phosphorylate DesR (13Albanesi D. Mansilla M.C. de Mendoza D. J. Bacteriol. 2004; 186: 2655-2663Crossref PubMed Scopus (84) Google Scholar) so we suggested that this covalent modification of DesR allows its interaction with the des regulatory region, leading to activation of des transcription. This assumption was based on the fact that no bacterial response regulator has yet been found to be active in the unphosphorylated form, although some response regulators, such as yeast SSK1 I, are inactive in the phosphorylated form (14Posas F. Saito H. EMBO J. 1998; 17: 1385-1394Crossref PubMed Scopus (249) Google Scholar). Although genetic studies have shown that DesR is absolutely necessary for des expression, it is unknown whether DesR is able to stimulate transcription per se or it is required to stimulate the synthesis or activity of another factor necessary for des transcription. In the present study we have examined how phosphorylation influences the ability of DesR to bind to the des promoter (Pdes). 1The abbreviations used are: Pdes, des promoter; RNAP, RNA polymerase; EMSA, electrophoretic mobility shift assay; Ac-P, acetyl phosphate; CI and CII, complexes I and II; RA and RB, regions A and B; DR, direct repeat. 1The abbreviations used are: Pdes, des promoter; RNAP, RNA polymerase; EMSA, electrophoretic mobility shift assay; Ac-P, acetyl phosphate; CI and CII, complexes I and II; RA and RB, regions A and B; DR, direct repeat. Our results show that only the phosphorylated form of DesR, DesR-P, is able to bind to the regulatory region it controls. In addition, we demonstrate that DesR is a dimer in solution that tends to tetramerize upon phosphorylation, which could account for cooperative interaction between molecules of DesR-P bound at adjacent DesR regulatory binding sites at Pdes. We show that DesR-P activates des transcription through recruitment of the RNA polymerase (RNAP) to the des promoter. A model for DesR-P promoter recognition has emerged from our results that we discuss in the context of transcriptional regulation of membrane fluidity homeostasis. Bacterial Strains, Plasmid, and Strain Constructions—The bacterial strains used in this work are listed in Supplementary Materials Table I. β-Galactosidase was assayed as previously described (15Mansilla C. de Mendoza D. J. Bacteriol. 1997; 179: 76-981Crossref Google Scholar). To mutate the IR-L inverted repeat of the regions that were protected in footprinting experiments (Fig. 4), oligonucleotides I, II, III, and IV were used for the amplification of two overlap-extension PCR (see Supplementary Materials Table II). A mixture of these PCR products was used as DNA template for another PCR using oligonucleotides I and VII (16Horton R. Hunt H. Ho S. Pullen J. Pease L. Gene (Amst.). 1989; 77: 61-68Crossref PubMed Scopus (2634) Google Scholar). The 301-bp amplification product was cloned into the integrational vector pJM116 (17Dartois V. Dérbarbouillé F. Kunst F. Rapaport G. J. Bacteriol. 1998; 180: 1855-1861Crossref PubMed Google Scholar) to generate pLC118. To mutate IR-R, oligonucleotides V and VI were utilized instead of II and III, and the resulting plasmid was pLC119. These plasmids were introduced into the amyE locus of strain JH642 giving strains LC118 and LC119, respectively. The same strategy was used to mutate the putative phosphorylation site Asp-54 with Ala, using oligonucleotides VIII, IX, X, and XI. The amplification product was cloned into plasmid pGS791 2G. Schujman, personal communication. under the xylose promoter, yielding plasmid pLC90. This plasmid was integrated ectopically in the thrC locus of a B. subtilis strain that lacks DesR and contains a fusion of the lacZ reporter gene to the des promoter (AKP9, see Ref. 6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar) yielding strain LC90. Wild type desR was amplified using oligonucleotides VIII and XI, and was cloned in the same way as desRD54A, giving strain LC89. All plasmids were sequenced to confirm the introduction of mutations. Gel Mobility Shift Assays—DNA fragments including the wild type or different des promoter variants were prepared by PCR using primers XIII(PECO895) and VII(BAMO) (Supplementary Materials Table II). The assay was performed as described previously (6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar). When phosphorylated DesR was used in the reaction, 6 μm DesR was incubated for 20 min at 0 °C in a mixture containing 50 mm Tris-HCl, pH 8, 5 mm MgCl2, 0.5 mm EDTA, 1.25 mm dithiothreitol, 10% glycerol, and 50 mm acetyl phosphate. For the ternary complex formation, the DNA fragment was prepared using the primers XIV and XV (Supplementary Materials Table II). The RNAP Holoenzyme (Epicenter) was added at 100 nm and incubated 30 min at 25 °C before running the gel. When heparin was included in the reaction, it was added for 5 min, at 60 μg/ml, after the incubation with the RNAP. OH•Footprinting Assays—Hydroxyl radical treatment was performed essentially as described (18del Solar G.H. Pérez-Martin J. Espinosa M. J. Biol. Chem. 1990; 265: 12569-12575Abstract Full Text PDF PubMed Google Scholar). A typical reaction mixture contained 2.6 nm 5′-end-labeled DNA fragment extending from –167 to +10 relative to the des transcription start site, DesR-P (300 or 1000 nm), 25 mm Tris-HCl, pH 8, 1 mm dithiothreitol, 0.25 mm EDTA, 4 mm MgCl2, and 200 ng of poly(dI-dC). After 30 min incubation at room temperature, 1 mm sodium ascorbate, H2O2 to 0.03%, and Fe(EDTA)2– (final concentrations of 18 and 36 μm for Fe(II) and Na2EDTA, respectively) were added, and the incubation was continued for 1 min at the same temperature. The reactions were stopped by the addition of a thiourea/EDTA solution (final concentrations of 9.5 mm thiourea and 1.7 mm EDTA). Mixes were run in a non-denaturing 5% polyacrylamide gel. The gel was exposed to an autoradiographic film for 5 min, and complexes I and II were cut out of the gel. The DNA was eluted, precipitated, loaded on denaturing 8% polyacrylamide sequencing gels, and run together with the products of sequencing reactions of Maxam and Gilbert (G + A) obtained from the same labeled DNA. The band pattern was visualized using a Fuji Film Phosphorimager. Separation of DesR from DesR-P—DesR or DesR phosphorylated in vitro with acetyl phosphate (Ac-P) for 20 min were injected on a C4 reversed phase Vydac column, using an ÄKTA Liquid chromatograph. Eluent A was 0.1% trifluoroacetic acid, 20% acetonitrile in water; eluent B was 0.1% trifluoroacetic acid, 60% acetonitrile in water. The sample was eluted with a gradient that varied from 80 to 0% A, over 40 min. Eluent B started at 20% and went to 100% over the same interval. The flow rate was 1 ml/min. Analytical Ultracentrifugation—Analytical ultracentrifugation studies were performed in a Beckman XL-A (Beckman Instruments Inc.) equipped with a UV-visible optic system inside the microcentrifuge. Sedimentation equilibrium experiments were carried out at 12,000 rpm and 20 °C in Epon eight-channel centerpieces with protein loading concentrations in the range of 1–12 μm. DesR was equilibrated in a buffer containing 50 mm Tris-HCl, pH 8, 5 mm MgCl2, 0.5 mm EDTA, 0.1 mm dithiothreitol (and 50 mm Ac-P when indicated). Radial absorption scans at 231, 275, and 280 nm were taken after the sedimentation equilibrium was reached. Baseline offsets were determined by high-speed centrifugation (42,000 rpm). Experimental data were analyzed using the program Sednterp version 1.03 XLAEQ and EQASSOC (Beckman; 19Laue T.M. Shah B.D. Ridgeway T.M. Pelletier S.L. Harding S.E. Rowe A.J. Horton J.C. Analytical Ultracentrifugation in Biochemistry and Polymer Science. Royal Society of Chemistry, Cambridge1992: 90-125Google Scholar, 20Minton A.P. Schuster T.M. Laue T.M. Modern Analytical Ultracentrifugation. Birkhauser, Boston, MA1994: 81-93Crossref Google Scholar), using a partial specific volume for DesR of 0.7449 ml g–1. This value was determined on the basis of its amino acid residue composition. The values of extinction coefficients at 280 nm were calculated from the known tryptophan and tyrosine residue contents. Determination of the Molecular Weight of DesR by Static Light Scattering—The average molecular weight (Mr) of wild type DesR and its mutant DesRD54A treated and non-treated with Ac-P were determined on a Precision Detector PD2010 light scattering instrument tandemly connected to an high performance liquid chromatography system and a LKB 2142 differential refractometer. The samples were loaded on a Sephadex G-25 (1 ml) column and eluted with 50 mm Tris-HCl, pH 8, 0.3 m NaCl, 0.5 mm dithiothreitol, at a flow rate of 0.3 ml/min. The 90° light scattering and refractive index signals of the eluting material were recorded on a PC computer and analyzed with the Discovery32 software supplied by Precision Detectors. The 90° light scattering detector was calibrated using bovine serum albumin (66.5 kDa) as a standard. In Vitro Transcription Assays—In vitro transcription reactions (50 μl) were performed in a buffer containing 40 mm Tris-HCl, pH 8, 10 nm MgCl2, 150 mm KCl, 0.1 mm EDTA, and 5% glycerol. A DNA fragment extending from position –177 to position +43 was prepared by PCR using oligonucleotides PA and PC (Supplementary Table II). The DNA (5 nm) was incubated first with DesR-P (300 nm) for 20 min at 25 °C, and then 10 min with RNAP (0.2 units) at the same temperature. The mixture was then treated with 10 μg/ml heparin for 3 min at 37 °C. Reactions were started by the addition of 5 μl of a solution 2 m of the four NTPs containing 0.5 μm [α-32P]UTP (400 Ci/mmol; 10 μCi/μl). After 5 min of incubation at 37 °C, the reaction was terminated by the addition of sodium acetate, pH 7, to 300 mm, EDTA to 15 mm, and tRNA to 100 μg/μl, and the nucleic acids were precipitated by the addition of 250 μl of 100% ethanol. The products were fractionated by electrophoresis on 8% polyacrylamide gels containing 8 m urea. Equation to Calculate the Relative Dissociation Constant of Regulator-DNA Complexes I and II—KdI and KdII represent the dissociation constants for specific complexes I and II, respectively. fmaxCI represents the fractional maximum level of complex I attained in the reactions (21Tsai S. Tsai M. O'Malley B. Cell. 1989; 57: 443-448Abstract Full Text PDF PubMed Scopus (194) Google Scholar). DNA+DesR⇄KdICI⇄KdII↘DesRCII(Eq. 1) KdIKdII=[1fmaxCI−1]2(Eq. 2) Phosphorylation of DesR Increases the Affinity of Binding to Its Target DNA—We wished to examine how phosphorylation influences the ability of DesR to bind to a DNA region spanning Pdes by electrophoretic mobility shift assays (EMSA) using DesR or DesR-P, this latter phosphorylated in vitro with Ac-P. To this end, a PCR-amplified [α-32P]dATP-labeled 216-bp DNA fragment extending from positions –187 to +29, relative to the des transcription start site, was incubated with purified DesR or DesR-P proteins. The results showed that Pdes exhibited changes in its electrophoretic mobility in the presence of either DesR (Fig. 1A) or DesR-P (Fig. 1B); however, when DesR-P was used, lower protein concentrations were sufficient to obtain the same fraction of complexed DNA. Phosphorylated DesR showed binding to the des promoter at concentrations as low as 0.06 μm. At this concentration a fast migrating complex I (CI) was mainly observed. Increasing DesR-P concentration enhanced progressively the formation of the slow-migrating complex II (CII; Fig. 1B). At 0.6 μm, CII is clearly the predominant protein-DNA complex. Formation of two different complexes, CI and CII, suggested to us the existence of two binding sites for the response regulator. DesR without Ac-P treatment was also able to bind to Pdes, but higher concentrations were needed to form CI and CII (Fig. 1A). DesR Phosphorylation Site Is Asp-54—Sequence comparative analysis of bacterial response regulators using multiple sequence alignment with hierarchical clustering (Multalin) indicated that Asp-54 of DesR should be the residue that receives the phosphoryl group from the cognate kinase, DesK. To determine whether this putative phosphorylation site, Asp-54, was involved in DesR activity, it was substituted with Ala using site-directed mutagenesis. DNA fragments carrying either desR+ (wild type) or desR encoding the mutant protein (desRD54A) were provided in trans into a desR null mutant carrying a fusion of the lacZ reporter gene to the des promoter. The expression of wild type desR+ allowed stimulation of des transcription upon a temperature downshift, whereas the expression of desRD54A could not stimulate transcription of the des-lacZ fusion under the same conditions (Supplemental Materials Fig. S1), indicating that residue Asp-54 is crucial for in vivo DesR transcriptional activity. Confirming this conclusion, we have determined that autophosphorylated DesK serves as a phosphodonor of the effector protein DesR, but cannot transfer the phosphoryl group to DesRD54A (data not shown). The Unphosphorylated Form of DesR Is Unable to Bind to the des Promoter—Some response regulators such as B. subtilis PhoP bind to its target promoters either in the unphosphorylated or in the phosphorylated form. However, the unphosphorylated form of PhoP is unable to activate transcription (22Liu W. Hulett M. J. Bacteriol. 1997; 179: 6302-6310Crossref PubMed Google Scholar). As shown in Fig. 1A, DesR binds to Pdes even without previous treatment with Ac-P, although in vivo experiments (described in Supplemental Materials Fig. S1) indicated that phosphorylation of Asp-54 is crucial for DesR activity. This suggested that DesR behaved similarly to PhoP. If this were the case, DesRD54A should bind to Pdes in EMSA, in the same way as unphosphorylated DesR, but treatment of DesRD54A with Ac-P should not increase its binding affinity. To evaluate this possibility, binding of purified DesRD54A to Pdes was tested using EMSA. Surprisingly, DesRD54A was unable to bind to Pdes at any tested concentration even when treated with Ac-P (Supplemental Materials Fig. S2A). This result could be attributed to: (i) the unphosphorylated form of DesR does not bind to Pdes, and the binding activity observed in the experiment shown in Fig. 1A was because of nonspecific phosphorylation of DesR during its overproduction in Escherichia coli, or (ii) the point mutation Asp-Ala in DesRD54A provoked a conformational change that abolished the binding of the mutated response regulator to Pdes. To distinguish between these two possibilities, DesR purified from E. coli was subjected to reverse phase high performance liquid chromatography (Supplemental Materials Fig. S2B). DesR showed two peaks that correspond to the phosphorylated and unphosphorylated form of the regulator. When DesR extracted from E. coli was incubated with Ac-P, the peak corresponding to DesR-P increased considerably (Supplemental Materials Fig. S2C). This result indicates that about 50% of His-tagged DesR purified from E. coli is phosphorylated, and that this form could be responsible for the binding activity of DesR without Ac-P treatment in the EMSA assays showed in Fig. 1A. It is worth mentioning that circular dichroism analysis showed that DesRD54A presents the same global secondary structure as wild type DesR (Supplemental Materials Fig. S2D), supporting the conclusion that this mutant regulator does not bind to Pdes because it is unable to be phosphorylated in the conserved aspartate by Ac-P, rather than by a structural change produced by the mutation. To demonstrate directly that the unphosphorylated form of DesR does not bind to its target promoter, we took advantage of the observation that DesR-P is completely dephosphorylated after incubation at 25 °C for 12 h (13Albanesi D. Mansilla M.C. de Mendoza D. J. Bacteriol. 2004; 186: 2655-2663Crossref PubMed Scopus (84) Google Scholar). We tested by EMSA the retardation of Pdes in the presence of untreated DesR (as it elutes from the nickel-nitrilotriacetic acid column), DesR dephosphorylated by incubation at 25 °C for 12 h, or DesR dephosphorylated at 25 °C and phosphorylated again by treatment with Ac-P. Fig. 2A shows that untreated DesR forms two complexes (lanes 2–4), whereas DesR dephosphorylated at 25 °C is unable to bind to Pdes (lanes 5–7). As expected, when this unphosphorylated DesR was incubated with Ac-P, it recovered the DNA binding activity (lanes 8–10). From now on, we will call "unphosphorylated DesR" the protein that was incubated 12 h at 25 °C. In addition, incubation of unphosphorylated DesR with increasing concentrations of Ac-P resulted in a similar enhancement of the binding of the response regulator, reaching a maximum at 50 mm of the phosphoryl donor (Fig. 2B). From the above results we can draw two conclusions: (i) the extent of phosphorylation of DesR determines its binding capacity, and (ii) unphosphorylated DesR is unable to bind to its target promoter. It should be mentioned that when DesRD54A was incubated with 50 mm Ac-P, and then subjected to high performance liquid chromatography, we detected that a secondary site(s) of the protein was phosphorylated (data not shown). Nevertheless, as shown in Supplemental Materials Fig. S2A, phosphorylation of this secondary site was unable to stimulate DesRD54A binding to Pdes. DesR Is a Dimer in Solution and Tetramerizes upon Phosphorylation—Phosphorylation promotes the formation of dimers or even oligomers of many response regulators; however, other response regulators do not change their association state upon phosphorylation (23Jeon Y. Lee Y. Han J. Kim J. Hwang D. J. Biol. Chem. 2001; 276: 40873-40879Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar, 24Lewis R. Scott D. Brannigan J. Ladds J. Cervin M. Spiegelman G. Hogget J. Barák I. Wilkinson A. J. Mol. Biol. 2002; 316: 235-245Crossref PubMed Scopus (1) Google Scholar, 25Wyman C. Rombel I. North A. Bustamente C. Kustu S. Science. 2000; 275: 1658-1661Crossref Scopus (209) Google Scholar, 26Da Re S. Schumacher J. Rousseau P. Fourment J. Ebel C. Kahn D. Mol. Microbiol. 1999; 34: 504-511Crossref PubMed Scopus (89) Google Scholar). The existence of two DesR-P·DNA complexes (Figs. 1 and 2) indicates that phosphorylation of DesR favors the binding of more than one subunit of the protein to the des promoter. To study the association state of DesR or DesR-P, we performed sedimentation equilibrium experiments by analytical ultracentrifugation. In the case of DesR-P it was necessary to define the required experimental conditions to stabilize the phosphorylated form during the course of the sedimentation assay. Because the half-life of DesR-P at 25 °C is about 90 min (13Albanesi D. Mansilla M.C. de Mendoza D. J. Bacteriol. 2004; 186: 2655-2663Crossref PubMed Scopus (84) Google Scholar), a high steady state of phosphorylation was maintained in situ by the inclusion of an excess of Ac-P in the sample subjected to analytical ultracentrifugation. Ultracentrifugation runs were performed either with unphosphorylated DesR or with DesR-P at concentrations in the range of 1 to 12 μm. Unphosphorylated DesR behaved as a dimer in solution at every tested concentration, because all the experimental data fit the Sednterp predicted curve for a dimer (Supplemental Materials Fig. S3A). When DesR was subjected to analytical ultracentrifugation in the presence of Ac-P, its hydrodynamic behavior changed considerably, especially at the highest protein concentration, at which the tetramer became the most abundant species (Supplemental Materials Fig. S3B), although the best fit includes constants for higher order multimers. This suggests that phosphorylation of DesR promotes new interaction between DesR dimers. Plots of the buoyant molecular mass versus protein concentration (1–12 μm) indicated that at low protein concentrations, the molecular mass of DesR-P corresponded to a dimer (Supplemental Materials Fig. S3C). As the levels of DesR-P increased, the mass of the species became larger, reaching the mass of a tetramer. This is a characteristic behavior of a system in equilibrium; otherwise dilution would not affect the average molecular mass of the species. To confirm the hydrodynamic properties of DesR obtained by analytical ultracentrifugation and to investigate the tetramerization phenomenon in the absence of continuous phosphorylation by Ac-P, we resort to static light scattering experiments. To this end, solutions of unphosphorylated DesR or DesR incubated with Ac-P were dialyzed for 20 min to remove the phosphorylating agent. Then, the total scattered intensity at 90° was examined. The average molecular weight of DesR was 41,553, and that of DesR-P was 71,834, corresponding to 1.94 and 3.36 units of DesR, respectively. The latter fractional number suggested that in this condition phosphorylation with Ac-P does not reach completion, and that about 70% of DesR was phosphorylated on this sample. Control experiments with DesRD54A showed an average molecular weight corresponding to a dimer regardless of the Ac-P treatment. Taken together, these experiments demonstrated that DesR is a dimer in solution that mainly forms tetramers upon phosphorylation. High-resolution Contacts of DesR on Its Target—Previous DNase I footprinting assays have shown that DesR binds to a DNA sequence that extends from positions –28 to –77 relative to the des transcription start site, generating 5 protected regions on each strand (6Aguilar P.S. Hernandez-Arriaga A.M. Cybulski L.E. Erazo A.C. de Mendoza D. EMBO J. 2001; 20: 1681-1691Crossref PubMed Scopus (304) Google Scholar). Two inverted repeats (5′-TCAT-3′) separated by 9 nucleotides were found in the center of the protected region. To obtain a high resolution profile of the contacts between DesR-P and its target DNA we resorted to