SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

The ‘two-component’ transcriptional activator FixJ controls nitrogen fixation in Sinorhizobium meliloti. Phosphorylation of FixJ induces its dimerization, as evidenced by gel permeation chromatography and equilibrium sedimentation analysis. Phosphorylation-induced dimerization is an intrinsic property of the isolated receiver domain FixJN. Accordingly, chemical phosphorylation of both FixJ and FixJN are second-order reactions with respect to protein concentration. However, the second-order phosphorylation constant is 44-fold higher for FixJN than for FixJ. Therefore, the C-terminal transcriptional activator domain FixJC inhibits the chemical phosphorylation of the receiver domain FixJN. Conversely, FixJN has been shown previously to inhibit FixJC activity ≈ 40-fold, reflecting the interaction between FixJN and FixJC. Therefore, we propose that modulation of FixJ activity involves both its dimerization and the disruption of the interface between FixJN and FixJC, resulting in the opening of the protein structure. Alanine scanning mutagenesis of FixJN indicated that the FixJ~P dimerization interface involves Val-91 and Lys-95 in helix α4. Dimerization was required for high-affinity binding to fixK promoter DNA.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

The expression of symbiotic nitrogen fixation genes of the soil bacterium Sinorhizobium meliloti undergoes a cascade regulation controlled by the FixLJ two-component regulatory system (David et al., 1988). FixLJ follows the canonical two-component model: the haem-containing FixL histidine protein kinase senses environmental oxygen (De Philip et al., 1990; Gilles-Gonzalez et al., 1991) and phosphorylates FixJ on a conserved aspartate residue (Lois et al., 1993; Reyrat et al., 1993). Phosphorylated FixJ (FixJ~P) activates transcription of the target genes nifA and fixK, whereas native non-phosphorylated FixJ is essentially inactive (Agron et al., 1993; Reyrat et al., 1993). FixJ shares with other two-component response regulators a typical two-domain modular arrangement (Parkinson and Kofoid, 1992; Stock et al., 1995), with a phosphorylatable N-terminal ‘receiver’ domain (FixJN) and a C-terminal transcriptional activator domain (FixJC). Within the native non-phosphorylated protein, the FixJN receiver domain negatively regulates the activity of the FixJC effector domain for both DNA binding and transcriptional activation (Kahn and Ditta, 1991; Da Re et al., 1994; Galinier et al., 1994).

Phosphorylation of FixJN relieves this inhibition in a manner that is not yet fully understood. Two classes of models have been proposed to describe how the phosphorylation of two-component regulators could activate a response. In one class of models, the receiver domain acts as an inhibitory module within the response regulator, and phosphorylation triggers a conformational change in the receiver domain, which abolishes this inhibition. Evidence for such a model has been provided for FixJ (Kahn and Ditta, 1991; Da Re et al., 1994), DctD (Lee et al., 1994), Spo0A (Ireton et al., 1993) and CheB (Simms et al., 1985). In the last case, the CheB crystallographic structure suggests that the receiver domain obstructs the methylesterase active site on the output domain (Djordjevic et al., 1998). In the structure of the FixJ homologue NarL, steric interference of the receiver domain prevents DNA binding via the C-terminal domain helix–turn–helix motif, implying a negative mode for regulation of the output domain activity by the receiver domain (Baikalov et al., 1996). Therefore, the structure must open up after phosphorylation to allow for DNA binding. In the second class of models, phosphorylation of the receiver domain activates the response regulator by triggering its oligomerization. For instance, the NtrC dimer has been shown to oligomerize upon phosphorylation, which is essential for its activation (Porter et al., 1993). Another example is the OmpR protein which, when phosphorylated co-operatively, binds to ompF promoter DNA (Huang et al., 1997). Here, we will show that these two classes of models are not mutually exclusive in the case of FixJ.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Phosphorylation-induced FixJ dimerization

In order to characterize the phosphorylated form of FixJ (FixJ~P), we phosphorylated FixJ chemically with acetyl-phosphate and separated FixJ~P from FixJ by ion exchange chromatography (Fig. 1A). In vitro transcription experiments on the nifA promoter confirmed that the second peak corresponds to the active phosphorylated form FixJ~P, whereas the first peak corresponding to native FixJ was inactive (data not shown). Both purified fractions were analysed by gel permeation chromatography. The FixJ fraction eluted from a G3000-SW-XL column as a single peak with an apparent molecular weight of 28 kDa (Fig. 1B, dotted line). The FixJ~P fraction eluted with an apparent molecular weight of 47 kDa, suggesting a dimeric structure (Fig. 1B, solid line). The dimeric nature of FixJ~P was confirmed by analytical sedimentation equilibrium experiments (Fig. 2). In these experiments, the FixJ molecular weight was estimated at 23.7 ± 1.8 kDa, close to the theoretical value of 22.2 kDa for a monomer. The sedimentation profile showed no detectable tendency towards dimerization of non-phosphorylated FixJ. Consistent with this, sedimentation velocity was found to be identical at 0.6 mg FixJ ml−1 and at 1.5 mg ml−1 (s = 2.2 S; data not shown). In order to study FixJ~P, it was necessary first to define experimental conditions stabilizing the phosphorylated form during the course of the sedimentation experiment. Indeed, the aspartyl-phosphate derivative of two-component regulators is usually quite labile. However, FixJ~P is unusual in this respect, with a rather long half-life of ≈ 2 h in the presence of Mg2+ and an even higher stability in the presence of EDTA (Weinstein et al., 1993). A preliminary study of the stability of the phosphorylated form indicated that it could be stabilized by an increase in pH and ionic strength. This allowed us to define the following conditions suitable for sedimentation experiments. FixJ~P was purified by ion exchange chromatography and dialysed against 50 mM Tris-HCl, pH 8, 0.1 mM EDTA, 0.1 mM dithiothreitol (DTT), 200 mM NaCl, before sedimentation analysis. Under these conditions, FixJ~P was found to be completely stable for at least 72 h when kept at 4°C. The large increase in the sedimentation coefficient of the FixJ~P sample, when compared with FixJ (2.6 S for FixJ~P concentrations between 0.18 and 0.75 mg ml−1), shows that, when phosphorylated, FixJ multimerizes rather than undergoes a conformational change (the increase in the hydrodynamic radius observed from gel filtration would lead to a decreased value of the sedimentation coefficient if the protein remained monomeric). When sedimentation equilibrium profiles were analysed as for a single molecular species, the apparent molecular weight of FixJ~P was estimated at 40 ± 3 kDa, slightly below the theoretical value of 44.4 kDa for a dimer. Because the FixJ~P preparation was known to be contaminated with non-phosphorylated FixJ, the sedimentation data were analysed further as for two molecular species, FixJ~P and FixJ, with the known Mr = 22.2 kDa for FixJ. The data were fitted by least square regression analysis with two parameters: the amount of monomeric FixJ and the molecular weight of FixJ~P. The best fit was obtained for an estimated molecular weight of 43 ± 3 kDa for FixJ~P, with amounts of FixJ~P ranging from 79% to 88% at the top and bottom of the sedimentation cell respectively (Fig. 2). We conclude from these experiments that the FixJ molecule dimerizes upon phosphorylation. Phosphorylation-induced dimerization has also been shown directly for two other response regulators, Spo0A (Asayama et al., 1995) and PhoB (McCleary, 1996).

image

Figure 1. . Phosphorylated FixJ is a dimer. A. Separation of FixJ~P by DEAE ion exchange chromatography. FixJ (140 μg) was preincubated (solid line) or not (dotted line) with acetyl-phosphate before separation. OD215 tracing, 0.25–0.45 M NaCl gradient. B. Analytical gel permeation of FixJ~P. Purified fractions from (A) were loaded on a G3000-SW column. The solid line corresponds to FixJ~P (apparent Mr = 47 kDa), whereas the dotted line corresponds to FixJ (apparent Mr = 28 kDa). OD215 tracing.

Download figure to PowerPoint

image

Figure 2. . Sedimentation equilibrium analysis of FixJ and FixJ~P. Centrifugation was carried out at 15 000 r.p.m. and at 20°C for 24 h. Bottom. Sedimentation profiles of FixJ and FixJ~P (left and right respectively). Top. Residuals expressed as the differences between experimental and fitted values. The FixJ sedimentation data were fitted as for a single molecular species with an estimated Mr = 23.7 ± 1.8 kDa. The FixJ~P sedimentation data were fitted as for two molecular species, FixJ~P and FixJ (Mr = 22.2 kDa), with an estimated Mr = 43 ± 3 kDa for FixJ~P and an average relative amount of 85% FixJ~P. For comparison, the dotted line shows the best fit for a monomeric species only (Mr = 22.2 kDa).

Download figure to PowerPoint

As the phosphorylation reaction produces dimers from FixJ monomers, it is expected to be second order with respect to FixJ concentration. We therefore investigated the kinetic order of the phosphorylation reaction with respect to FixJ. Increasing concentrations of FixJ were phosphorylated in the presence of acetyl-phosphate. Phosphorylation reactions were quenched on ice with an excess of EDTA, then analysed by gel permeation chromatography to allow for quantification of FixJ~P. This analysis showed that FixJ phosphorylation is a second-order reaction with respect to FixJ concentration (Fig. 3), consistent with the experiments above showing the dimeric nature of FixJ~P.

image

Figure 3. . Second-order phosphorylation of FixJ and FixJN. FixJ (circles) or FixJN (squares) was phosphorylated at various protein concentrations in the presence of acetyl-phosphate for 80 min at 25°C. The molarities of the phosphorylated species are plotted against the squares of the molarities of the non-phosphorylated species.

Download figure to PowerPoint

Phosphorylation-induced dimerization of the FixJ receiver domain

In order to study the basis of phosphorylation-induced FixJ dimerization, we dissected FixJ into its two domains, the FixJN receiver domain and the FixJC transcriptional activator domain (Da Re et al., 1994). A FixJN-expressing plasmid, pGMI1955, was engineered to overexpress the first 126 amino acids of FixJ, and the FixJN protein was purified by ion exchange and gel permeation chromatography. Like FixJ (Agron et al., 1993; Reyrat et al., 1993), FixJN could be phosphorylated in vitro by the cognate FixL histidine kinase (Fig. 4). Thus, the FixJN receiver domain can be phosphorylated autonomously in the absence of the FixJ C-terminal domain and functions as a distinct module within the FixJ molecule.

image

Figure 4. . FixL-dependent phosphorylation of FixJ and FixJN. Proteins were incubated as indicated (8.8 μM FixL, 5.2 μM FixJ, 5.8 μM FixJN) in the presence of [γ-32P]-ATP for 40 min at 28°C. Autoradiogram of a 20% acrylamide denaturing SDS gel.

Download figure to PowerPoint

Gel permeation was used to analyse the quaternary structure of FixJN after phosphorylation. When FixJN was phosphorylated in the presence of acetyl-phosphate, a new species appeared with a higher molecular weight (Fig. 5). To confirm that this new species is indeed phosphorylated FixJN, we used [33P]-acetyl-phosphate and found that the label associated with the high-molecular-weight peak (data not shown). FixJN phosphorylation was found to be reversible in the presence of Mg2+, with an estimated half-life for FixJN~P of 73 min at 25°C in TED buffer, pH 7.2, containing 0.3 M KCl and 10 mM MgCl2. In a control experiment, the C-terminal domain FixJC was treated with acetyl-phosphate and was found to be monomeric independently of acetyl-phosphate. Therefore, dimerization of FixJ can be attributed to the autonomous dimerization of the FixJN receiver domain upon phosphorylation.

image

Figure 5. . Phosphorylation-induced dimerization of FixJN. Analytical gel permeation chromatography on a TSK G3000-SW-XL column of (A) FixJN or (B) FixJ, before (solid line) or after (dotted line) treatment with acetyl-phosphate. OD215 tracing.

Download figure to PowerPoint

Interaction between FixJN and FixJC within the FixJ molecule

The kinetic order of FixJN phosphorylation was determined in a titration experiment performed as above, which showed that phosphorylation is a second-order reaction with respect to FixJN concentration (Fig. 3), in agreement with the dimeric structure of the phosphorylated protein. This experiment also showed that FixJN phosphorylates much more readily from acetyl-phosphate than FixJ: the second-order phosphorylation constant was 0.17 μM−1 for FixJN versus 0.004 μM−1 for FixJ. We conclude from these experiments that the FixJC domain inhibits the chemical phosphorylation of the FixJN receiver domain within the FixJ molecule.

This 44-fold inhibition of FixJN phosphorylation by FixJC can be interpreted in the light of the 40-fold inhibition of FixJC activity by FixJN that was observed previously (Da Re et al., 1994). We propose that this mutual inhibition is the direct consequence of an interaction between the FixJ receiver and output domains. The 40-fold mutual inhibition corresponds to a relatively weak interaction energy ΔG = RT ln40 = 2.2 kcal mol−1 between the two domains within the FixJ molecule. With such a weak interaction between the two domains, we predict that native FixJ protein exists in an equilibrium between a major closed inactive form and a minor open active form. The latter transcriptionally active form would also be more readily phosphorylated by acetyl-phosphate, like FixJN. In this model, only the open form, which represents ≈ 2.5% of the total protein, would be active in transcription, which accounts for the earlier observation that native, non-phosphorylated FixJ can activate transcription only at high protein concentration (Reyrat et al., 1993; Da Re et al., 1994). The weak interaction energy between FixJN and FixJC thus appears to be functionally relevant for the activation of FixJ, because it permits the opening of the structure at a moderate cost upon phosphorylation.

Mapping of the dimerization interface by alanine scanning mutagenesis

In order to map the dimerization interface, we performed alanine scanning mutagenesis of the FixJ receiver domain. Solvent-exposed residues were determined on the basis of the homologous structure of CheY (Stock et al., 1989), and the following sets were mutated to alanine: 1 {T2, D3}; 2 {E12, E13, K17, S18}; 3 {F21, M25}; 4 {K31, Q34}; 5 {F42, D45, R47}; 6 {R56, D59}; 7 {E64, R67}; 8 {D71, L72}; 9 {H84, G85, V87}; 10 {V91, K95, F101}; 11 {E113, R117, E120, V121}. The mutated proteins were purified as GST fusion proteins, the GST moiety was cleaved away with thrombin and the mutated FixJ proteins separated by ion exchange chromatography. For an unknown reason, the GST–FixJ7 fusion protein did not bind to the glutathione–Sepharose column and was not studied further. The other variant FixJ proteins were tested for acetyl-phosphate-dependent phosphorylation and dimerization. Mutated proteins FixJ2, FixJ3, FixJ4, FixJ5, FixJ8 and FixJ11 were phosphorylated and dimerized as efficiently as wild-type protein. The FixJ6 protein was phosphorylated very inefficiently, yet dimerized after phosphorylation; note that the mutations in FixJ6 affect the direct vicinity of the Asp-54 phosphorylation site. Other mutated proteins affected for phosphorylation are FixJ1, FixJ9 and FixJ10. The FixJ1 protein was slightly affected, and it dimerized upon phosphorylation with a second-order constant of 0.001 μM−1. The FixJ9 and FixJ10 proteins appeared not to be phosphorylated at all in the presence of acetyl-phosphate. This was verified both by direct isoelectric focusing and by the absence of labelling with [33P]-acetyl-phosphate. The triple mutation in fixJ10 was resolved further to single mutations and the mutated proteins purified as above. Only the F-101A mutation resulted in a complete loss of phosphorylation with acetyl-phosphate. Although the V-91A and K-95A mutations allowed FixJ phosphorylation, the phosphorylated proteins remained fully monomeric, showing that these residues are involved in the phosphorylation-induced dimerization of FixJ. These results map the dimerization interface to helix α4 of the receiver domain.

Dual effect of FixJ phosphorylation and the function of FixJ~P dimerization

Based on a comparison of the activities of the isolated C-terminal domain FixJC with full-length FixJ protein, we have shown previously that inhibition of the latent activity of FixJC by FixJN is relieved upon phosphorylation (Kahn and Ditta, 1991; Da Re et al., 1994). This model can be interpreted on the basis of the three-dimensional structure of the homologous NarL protein, in which the output domain homologous to FixJC is sterically hindered by the receiver domain (Baikalov et al., 1996). Thus, in both FixJ and NarL, we are led to postulate that the structure opens up upon phosphorylation in order to release the output domain for DNA binding and transcriptional activation. Therefore, phosphorylation of FixJ triggers a dual effect, both opening up of the FixJ structure and inducing FixJ dimerization. A similar dual effect of phosphorylation on the structure of a response regulator can also be suggested in the case of Spo0A, as the receiver domain inhibits the output domain (Ireton et al., 1993), while the protein dimerizes upon phosphorylation (Asayama et al., 1995).

What is the function of the dimerization of FixJ~P? Dimerization is not a prerequisite for transcriptional activation, as the nifA promoter can be activated by the isolated FixJC domain, which is a monomer (Da Re et al., 1994) and, to a lesser extent, by native FixJ, which is also monomeric (Fig. 2). Therefore, prior dimerization is not essential for transcriptional activation, although we cannot rule out the possibility that a functional promoter may require the simultaneous binding of two FixJ or FixJC protomers for activation. The major role for FixJ~P dimerization is likely to be to enhance its affinity and specificity for target sites such as the fixK promoter. In order to test this hypothesis, the non-dimerizable mutated proteins described above (V-91A and K-95A) were used in bandshift titration experiments with fixK promoter DNA (Fig. 6). This experiment showed that dimerization of FixJ~P contributes strongly (≈ 10-fold) to the affinity of FixJ~P for the fixK promoter. Similarly, the much higher affinity of FixJ~P compared with FixJC (Galinier et al., 1994) can be attributed to the difference in quaternary structure shown in the present work. We conclude that phosphorylation of the FixJ receiver domain triggers two different effects, both important for function. Opening up of the FixJ molecule is essential for activity, while dimerization contributes to DNA affinity. Thus, the two main classes of models for the activation of response regulators, relief of inhibition versus oligomerization, are not mutually exclusive.

image

Figure 6. . Contribution of FixJ~P dimerization to DNA binding. Bandshift assay titration of fixK promoter DNA with FixJ~P (A) and with the non-dimerizable FixJV91A~P (B). Increasing protein molarities were 1.25 μM, 2.5 μM, 5 μM, 10 μM, 20 μM, 30 μM and 60 μM (lanes 1–7).

Download figure to PowerPoint

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Plasmid constructions

Plasmid pGMI1955 used for FixJN overexpression was obtained as follows. The starting material was the pBluescript derivative pDK383, a plasmid expressing both FixJN and FixJC (Kahn and Ditta, 1991). The 0.39 kb EcoRV fragment containing fixJC was excised to generate pGMI1930, from which the fixJN sequence was amplified in order to engineer an NdeI site overlapping the initiation codon. pGMI1955 was obtained by cloning the 0.4 kb NdeI–EcoRV fragment between the NdeI and SmaI sites of pT7-7, thus bringing fixJN under the control of the T7 expression system.

For alanine scanning mutagenesis, 11 separate sets of mutations were introduced in vitro into fixJ using the Sculptor mutagenesis system (Amersham). The template plasmid was pJS2, a derivative of pDK302 (Kahn and Ditta, 1991) carrying an NdeI restriction site overlapping the fixJ initiation codon and a silent SacII restriction site between fixJN and fixJC. pJS2 derivatives were cut with NdeI and treated with T4 DNA polymerase before EcoRI cleavage. The isolated 0.88 kb fragments, containing the mutated fixJ coding sequences, were cloned between the SmaI and EcoRI sites of pGEX-2T (Pharmacia). The resulting plasmids allow for overexpression of the mutated FixJ proteins as GST fusions.

Protein purification

The FixJ and FixJC proteins were purified as described previously (Reyrat et al., 1993; Da Re et al., 1994 respectively). To purify FixJN, the E. coli strain BL21(DE3) containing pGMI1955 was grown at 37°C in LB medium in the presence of appropriate antibiotics. At mid-log exponential phase, expression of FixJN was induced for 4 h in the presence of IPTG (0.2 mM). Cells were harvested, resuspended in TED buffer (20 mM Tris-HCl, pH 7.2, 0.1 mM EDTA, 0.1 mM DTT) and disrupted by sonication. FixJN was fractionated between 40% and 60% ammonium sulphate saturation, dialysed against TD buffer (20 mM Tris-HCl, pH 7.2, 0.1 mM DTT), and loaded on a PorosHQ anion exchange column (7.5 × 75 mm; Perseptive Biosystems) equilibrated in the same buffer. FixJN was eluted with a linear gradient from 0.1 to 0.4 M NaCl (18 column volumes at 2 ml min−1). Fractions containing FixJN were pooled, concentrated and applied to a Sephacryl S200 column (2.2 × 82 cm; Pharmacia) equilibrated in TED buffer containing 0.1 M KCl. Fractions containing pure FixJN were pooled, concentrated with a Centripep-10 concentrator (Amicon) and dialysed against TED buffer containing 0.1 M KCl and 50% glycerol. The protein was at least 95% pure and was stored at −20°C.

Mutated GST–FixJ fusion proteins were overexpressed, purified and cleaved with thrombin according to the manufacturer's instructions (GST Gene Fusion System; Pharmacia). Thrombin cleavage was carried out directly on fusion proteins adsorbed on glutathione–Sepharose 4B. Cleaved proteins were separated from thrombin by anion exchange chromatography on Resource Q (1 ml; Pharmacia). FixJ was eluted with a 14 ml linear gradient from 0.15 M to 0.4 M NaCl in 50 mM Tris-HCl, pH 7.6, containing 0.1 mM DTT, then concentrated, dialysed and stored at −20°C in TED buffer containing 0.1 M KCl and 50% glycerol. Purified FixJ proteins were homogeneous as judged by SDS–PAGE analysis. The resulting FixJ proteins carry an amino-terminal Gly-Ser extension compared with FixJ purified by the conventional procedure.

Phosphorylation reactions

Acetyl-phosphate-dependent phosphorylation of FixJ or FixJN was achieved in 20 μl of 50 mM Tris-HCl, pH 7.2, 0.1 mM EDTA, 0.1 mM DTT, 10 mM MgCl2 and 20 mM acetyl-phosphate. After incubation for 80 min at 25°C, the reaction was stopped with 20 mM EDTA and transferred on ice. FixJ~P and FixJN~P were stable for several hours under these conditions. The amount of phosphorylated protein was determined by analytical gel permeation or ion exchange chromatography. FixL-dependent phosphorylation of FixJ was achieved as described by Reyrat et al. (1993).

Bandshift assays

FixJ or FixJV91A (both purified from GST fusion proteins) were phosphorylated as described above with acetyl-phosphate at a molarity of 100 μM protein. FixJ~P was serially diluted, incubated with [32P]-labelled pfixK DNA, and complexes were separated on 6% polyacrylamide gels as described previously (Galinier et al., 1994).

Separation of FixJ~P by ion exchange chromatography

FixJ (140 μg) was phosphorylated in the presence of acetyl-phosphate in a final volume of 12 μl. It was loaded onto an anion exchange column (TSK DEAE-5PW; 7.5 × 75 mm; Hewlett Packard) equilibrated in TD buffer and eluted with a linear gradient from 0.25 M to 0.45 M NaCl in TD. EDTA (1 mM) was added to each fraction in order to stabilize the phosphorylated form. Fractions of interest were analysed by gel permeation chromatography as described above. For preparative purposes, this procedure was scaled up to 7 mg of FixJ, which was phosphorylated as above in a final volume of 1.2 ml. The protein was loaded onto a Q HyperD 10 column (4.6 × 100 mm; Beckman) equilibrated in 20 mM Tris-HCl, pH 8, containing 0.1 mM EDTA and 0.1 mM DTT, and eluted with a linear gradient from 0.2 M to 0.5 M NaCl in the same buffer. FixJ~P-containing fractions were collected on EDTA as above, pooled and concentrated. The resulting FixJ protein was 93% phosphorylated as assessed by analytical anion exchange chromatography.

Analytical gel permeation chromatography

This was performed by high-performance liquid chromatography (HPLC) on a TSK G3000-SW-XL column (0.75 × 30 cm; Hewlett Packard) equilibrated in TED buffer containing 0.1 M KCl. Molecular weight standards were aproteinin (6.5 kDa), cytochrome c (12.4 kDa), carbonic anhydrase (29 kDa) and BSA (66 kDa).

Analytical ultracentrifugation

Sedimentation equilibrium experiments were performed using a Beckman model XL-A analytical ultracentrifuge. FixJ and FixJ~P were dialysed exhaustively against 50 mM Tris-HCl, pH 8, containing 0.1 mM EDTA, 0.1 mM DTT and 200 mM NaCl. The FixJ and FixJ~P sedimentation experiments were carried out at 20°C in 1.2 cm path length double-sector and hexa-sector cells, respectively, and run at 15 000 r.p.m. After centrifugation for 24 h, scans were compared at 2 h intervals to ensure that equilibrium had been reached, and an experimental baseline was determined by centrifugation at 42 000 r.p.m. Data were collected at 274 nm from which concentrations were derived on the basis of an estimated ε = 0.134 AU274 mg−1. Data were fitted by least square regression analysis using the following equation for each molecular species:

inline image

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

We wish to thank Jacqueline Cherfils for the calculation of solvent accessibility in CheY. This work was supported by grants from the European Union (BIO4 CT 97-2143), the Ministère de l'Education Nationale, de la Recherche et de la Technologie (Programme de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires), INRA (Programme prioritaire ‘Microbiologie’) and the Ministère des Affaires Etrangères (PROCOPE 97158).

Footnotes
  1. *Present address: Department of Molecular Biology, University of Princeton, Princeton, NJ 08544, USA

  2. †The first two authors contributed equally to this work

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  • 1
    Agron, P.G., Ditta, G.S., Helinski, D.R. (1993) Oxygen regulation of nifA transcription in vitro. Proc Natl Acad Sci USA 90: 35063510.
  • 2
    Asayama, M., Yakamoto, A., Kobayashi, Y. (1995) Dimer form of phosphorylated Spo0A, a transcriptional regulator, stimulates the Spo0F transcription at the initiation of sporulation in Bacillus subtilis. J Mol Biol 250: 1123.
  • 3
    Baikalov, I., Schröder, I., Kaczor-Grzeskowiak, M., Grzeskowiak, K., Gunsalus, R.P., Dickerson, R.E. (1996) Structure of the E. coli response regulator NarL. Biochemistry 35: 1105311061.
  • 4
    Da Re, S., Bertagnoli, S., Fourment, J., Reyrat, J.M., Kahn, D. (1994) Intramolecular signal transduction within the FixJ transcriptional activator: in vitro evidence for the inhibitory effect of the phosphorylatable regulatory domain. Nucleic Acids Res 22: 15551561.
  • 5
    David, M., Daveran, M.L., Batut, J., Dedieu, A., Domergue, O., Ghai, J., et al (1988) Cascade regulation of nif gene expression in Rhizobium meliloti. Cell 54: 671683.
  • 6
    De Philip, P., Batut, J., Boistard, P. (1990) Rhizobium meliloti FixL is an oxygen sensor and regulates R. meliloti nifA and fixK genes differently in Escherichia coli. J Bacteriol 172: 42554262.
  • 7
    Djordjevic, S., Gudreau, P.N., Xu, Q., Stock, A.M., West, A.H. (1998) Structural basis for methylesterase CheB regulation by a phosphorylation-activated domain. Proc Natl Acad Sci USA 95: 13811386.
  • 8
    Galinier, A., Garnerone, A.M., Reyrat, J.M., Kahn, D., Batut, J., Boistard, P. (1994) Phosphorylation of the Rhizobium meliloti FixJ protein induces its binding to a compound regulatory region at the fixK promoter. J Biol Chem 269: 2378423789.
  • 9
    Gilles-Gonzalez, M.A., Ditta, G.S., Helinski, D.R. (1991) A haemoprotein with kinase activity encoded by the oxygen sensor of Rhizobium meliloti. Nature 350: 170172.
  • 10
    Huang, K.-J., Lan, C.-Y., Igo, M. (1997) Phosphorylation stimulates the cooperative DNA-binding properties of the transcription factor OmpR. Proc Natl Acad Sci USA 94: 28282832.
  • 11
    Ireton, K., Rudner, D.Z., Siranosian, K.J., Grossman, A.D. (1993) Integration of multiple developmental signals in Bacillus subtilis through the Spo0A transcription factor. Genes Dev 7: 283294.
  • 12
    Kahn, D. & Ditta, G. (1991) Molecular structure of FixJ: homology of the transcriptional activator domain with the −35 binding site of sigma factors. Mol Microbiol 5: 987997.
  • 13
    Lee, J.H., Scholl, D., Nixon, B.T., Hoover, T.R. (1994) Constitutive ATP hydrolysis and transcription activation by a stable, truncated form of Rhizobium meliloti DCTD, a sigma 54-dependent transcriptional activator. J Biol Chem 269: 2040120409.
  • 14
    Lois, A.F., Weinstein, M., Ditta, G.S., Helinski, D.R. (1993) Autophosphorylation and phosphatase activities of the oxygen-sensing protein FixL of Rhizobium meliloti are coordinately regulated by oxygen. J Biol Chem 268: 43704375.
  • 15
    McCleary, W.R. (1996) The activation of PhoB by acetylphosphate. Mol Microbiol 20: 11551163.
  • 16
    Parkinson, J.S. & Kofoid, E.C. (1992) Communication modules in bacterial signalling proteins. Annu Rev Genet 26: 71112.
  • 17
    Porter, S.C., North, A.K., Wedel, A.B., Kustu, S. (1993) Oligomerisation of NTRC at the glnA enhancer is required for transcriptional activation. Genes Dev 7: 22582273.
  • 18
    Reyrat, J.M., David, M., Blonski, C., Boistard, P., Batut, J. (1993) Oxygen regulated in vitro transcription of Rhizobium meliloti nifA and fixK genes. J Bacteriol 175: 68676872.
  • 19
    Simms, S.A., Keane, M.G., Stock, J. (1985) Multiple forms of the CheB methylesterase in bacterial chemosensing. J Biol Chem 260: 1016110168.
  • 20
    Stock, A.M., Mottonen, J.M., Stock, J.B., Schutt, C.E. (1989) Three-dimensional structure of CheY, the response regulator of bacterial chemotaxis. Nature 337: 745749.
  • 21
    Stock, J.B., Surette, M.G., Levit, M., Park, P. (1995) Two-component signal transduction systems: structure–function relationships and mechanisms of catalysis. In Two-Component Signal Transduction. Hoch, J.A., and Silhavy, T.J. (eds). Washington, DC: American Society for Microbiology Press, pp. 2551.
  • 22
    Weinstein, M., Lois, A.F., Ditta, G.S., Helinski, D.R. (1993) Mutants of the two-component regulatory protein FixJ of Rhizobium meliloti that have increased activity at the nifA promoter. Gene 134: 145152.