The C-terminal receiver domain of the Rhizobium leguminosarum bv. viciae FixL protein is required for free-living microaerobic induction of the fnrN promoter

Bert Boesten and Ursula B. Priefer

Ökologie des Bodens, Botanisches Institut, RWTH-Aachen, Worringerweg 1, 52056 Aachen, Germany

Correspondence
Bert Boesten
boesten{at}rwth-aachen.de


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The Rhizobium leguminosarum bv. viciae VF39 FixL protein belongs to a distinct group of hybrid regulatory sensor proteins that bear a covalently linked C-terminal receiver domain. FixL has an unorthodox histidine kinase domain, which is shared with many other hybrid regulators. The purified FixL protein had autophosphorylation activity. A truncated protein, lacking the receiver domain, had a much-reduced autophosphorylation activity. However, this truncated protein still efficiently phosphorylated the purified receiver domain in trans. This indicates that, in the full-length FixL protein, the conserved histidine residue in the kinase domain is phosphorylated only transiently and that most of the phosphoryl label accumulates in the C-terminal receiver domain. Gene-fusion studies showed that the fixL gene is required for free-living microaerobic induction of the fnrN promoter. The presence of a functional fixK gene is not required. An R. leguminosarum strain lacking fixL could not be complemented with a truncated copy of the gene lacking the receiver domain. This indicates that the C-terminal receiver domain is an intermediate in the signal transduction pathway that links oxygen limitation to induction of the fnrN promoter in R. leguminosarum bv. viciae VF39.


Abbreviations: GST, glutathione S-transferase; GUS, {beta}-glucuronidase; HK, histidine kinase


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Many micro-organisms have to be able to cope with conditions of reduced oxygen pressure. In the Rhizobium–legume symbiosis, the fixation of atmospheric nitrogen by the microsymbiont takes place in a microaerobic environment. Reduced oxygen pressure is one of the physiological signals that lead to the coordinated expression of genes required for an efficient symbiosis. The paradigm of an oxygen-dependent regulatory cascade has been described for Sinorhizobium meliloti (David et al., 1988). At the top of this cascade stands the haem-containing FixL sensor protein and the FixJ transcriptional regulator protein. FixLJ belongs to a large family of two-component regulatory systems (Ronson et al., 1987). The sensor proteins of this family all share a central histidine kinase (HK) domain. This domain consists of an H box region containing a conserved histidine residue, which is the target of autophosphorylation, and a nucleotide-binding region (Parkinson & Kofoid, 1992). In FixL, a haem-containing PAS (period, ARNT and single-minded) sensory domain (Zhulin et al., 1997) is located immediately upstream from the HK domain. This haem domain is involved in sensing molecular oxygen and regulates the activity of the adjacent HK domain. In the absence of oxygen, FixL autophosphorylates and catalyses the phosphorylation of the transcriptional regulator FixJ (Gilles-Gonzalez et al., 1991; Tuckerman et al., 2001). The phosphorylated FixJ (FixJ-P) promotes the transcription of genes encoding two other transcriptional activators, NifA and FixK (Reyrat et al., 1993). These proteins in turn activate a range of genes involved in symbiotic nitrogen fixation.

Homologues of these regulatory genes and their target genes have been identified in many other rhizobia. From regulatory studies to date, it has become clear that, between the Rhizobium species, many differences exist in the interconnectivity between these regulatory genes and their target genes (Fischer, 1994). In Rhizobium leguminosarum bv. viciae VF39, genes encoding FixL and FixK homologues have also been identified (Patschkowski et al., 1996). The fixL gene is located immediately downstream from fixK, and the genes probably form an operon. The R. leguminosarum FixL protein is remarkable in that it has a C-terminal extension encoding a receiver domain. No gene encoding a FixJ homologue has been identified in R. leguminosarum bv. viciae VF39.

FixK belongs to the FNR/CRP superfamily of transcriptional regulators (Green et al., 2001). In S. meliloti, FixK is required for the activation of transcription of the fixNOQP operons. These genes code for a high-affinity cbbB3-type terminal oxidase and are essential for an efficient symbiosis (Preisig et al., 1993, 1996). In R. leguminosarum VF39, two copies of the fixNOQP operon exist. One copy is located on the c plasmid, just upstream of the fixK gene, and transcribed divergently. The second copy is located on the d plasmid. Despite the strong similarities with the S. meliloti FixK protein, which regulates the fixNOQP operons in that micro-organism, the R. leguminosarum VF39 FixK has been found to be only marginally involved in fixNOQP expression. Besides fixK, another gene encoding an FNR/CRP-type transcriptional regulator has been identified in R. leguminosarum bv. viciae VF39 (fnrN, Colonna-Romano et al., 1990). FnrN is found to be primarily responsible for activating the fixNc and fixNd promoters. The expression of the fixNOQP genes is also highly reduced in a fixL mutant background (Schlüter et al., 1997). This suggests that, in R. leguminosarum VF39, FixL may be involved in the regulation of FnrN.

In this work, we demonstrate that this is indeed the case. Furthermore, we show that the C-terminal receiver domain of the R. leguminosarum bv. viciae VF39 FixL protein is an intermediate in the phophoryl relay pathway involved in the oxygen-dependent regulation of the fnrN promoter.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains plasmids and media.
All bacterial strains and plasmids used in this work are listed in Table 1. Escherichia coli strains were grown at 37 °C in Luria–Bertani medium (Maniatis et al., 1982). Rhizobium leguminosarum strains were grown at 29 °C in TY complex medium (Beringer, 1974). Antibiotics used for E. coli were kanamycin (25 µg ml–1) or tetracycline (10 µg ml–1). Antibiotics used for R. leguminosarum were neomycin (25 µg ml–1), tetracycline (10 µg ml–1) or streptomycin (300 µg ml–1).


View this table:
[in this window]
[in a new window]
 
Table 1. Strains and plasmids used in this work

ApR, ampicillin resistance; CmR, chloramphenicol resistance; GmR, gentamicin resistance; KmR, kanamycin resistance; NmR, neomycin resistance; SmR, streptomycin resistance; TcR, tetracycline resistance.

 
Construction of an fnrN : : uidA gene fusion.
A transcriptional gene fusion of the fnrN promoter with the uidA {beta}-glucuronidase (GUS) reporter gene was constructed in the stable broad-host-range plasmid pJP2 (Prell et al., 2002). A 1·9 kb EcoRI–HinDIII fragment from pRlA76, carrying the fnrN gene from R. leguminosarum VF39 (Colonna-Romano et al., 1990), was cloned into pUC19. An 892 bp DraI–PstI DNA fragment containing the fnrN promoter region was then cloned into pK18mob and restricted with PstI and HinCII. Finally, the fnrN promoter fragment was excised from this plasmid as a BamHI–HinDIII fragment and cloned into pJP2 restricted with the same enzymes. The resulting plasmid was designated pMKJ-R.

Manipulations of the fixL gene.
A DNA fragment containing the entire sequence encoding FixL was amplified from VF39 chromosomal DNA by PCR with oligonucleotide primers 5 and 6 (Table 2; Fig. 4a). A BamHI restriction site just upstream of the ATG start codon and an EcoRI restriction site immediately downstream from the stop codon were incorporated in the oligonucleotide primers. The PCR product was digested with BamHI and EcoRI and cloned into the GST plasmid pGEX-5x-3 (Amersham Pharmacia Biotech). This resulted in plasmid pAS1, which encodes the FixLc fusion protein (Fig. 1). The BamHI–EcoRI fragment from pAS1 was cloned into pUC19 to give pJSC100. Point mutations were introduced in pJSC100 using the GeneEditor in vitro Site Directed Mutagenesis System (Promega) and mutagenic oligonucleotides 7 and 8 (Table 2; Fig. 4a). This resulted in the introduction of BamHI restriction sites between the PAS and the haem-binding domains (pJSC102) and between the HK domain and the C-terminal receiver domain (pJSC106). The truncated BamHI–EcoRI fragments were then excised from these plasmids and cloned into pGEX-5x-3. The resulting plasmids, pJSG103 and pJSG107, coded for the FixL3 and FixL7 fusion proteins, respectively (Fig. 1).


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotides used in this work

The location of binding of these oligonucleotides to the fixKL region is also indicated on the physical map at the top of Fig. 4. Bold type in primers 5 and 6 indicates the position of the introduced restriction sites. Bold type in primers 9 and 10 indicates the position of the TGA stop codon.

 


View larger version (21K):
[in this window]
[in a new window]
 
Fig. 4. Organization of the fixKLN region in VF39 wild-type and mutant strains. (a) Physical map of the region, showing relevant restriction sites and location of oligonucleotide binding sites. B, BstEII; H, HinDIII; K, KpnI; N, NotI; S, SalI. The numbering of the oligonucleotides corresponds with the numbers in Table 2. The solid arrowheads correspond to oligonucleotides that were used for PCR amplification. The open arrowheads correspond to mutagenic oligonucleotides. The arrowheads point in the direction of the extension reaction. (b) Genetic map of the fixKLN region in the wild-type R. leguminosarum bv. viciae VF39 strain. (c) A derivative of VF39 was constructed where the entire fixL gene and the C-terminal part of fixK were deleted from the genome (VF39{Delta}KL). (d) to (g) The VF39{Delta}KL strain was complemented with pK18mob derivatives containing various extents of the fixKL gene cluster. The plasmids were integrated into the VF39{Delta}KL genome by a single crossover event, indicated by the large ‘X’ in the fixKfixN promoter region. The resulting genotypes of the complemented strains are indicated on the right. KmR, kanamycin resistance.

 


View larger version (24K):
[in this window]
[in a new window]
 
Fig. 1. Domain organization of the FixL proteins. The S. meliloti FixL protein consists of an N-terminal transmembrane domain, a central haem domain and a histidine kinase (HK) domain. FixJ contains an N-terminal receiver domain and a transcriptional activator (output) domain. The R. leguminosarum FixL protein contains an N-terminal PAS domain, a haem domain, an HK domain and a covalently linked C-terminal receiver domain. There is no FixJ. The transmembrane domain anchors the S. meliloti FixL to the cytoplasmic membrane. The N-terminal domain of the R. leguminosarum FixL protein does not feature transmembrane domains. Instead, this domain contains a PAS fold. The haem domain also contains a PAS fold, which holds the haem prosthetic group and is involved in oxygen (O2) sensing. The HK domain features an ATP binding region and an H box motif. The ATP binding region consists of an N, G1, F and a G2 box, so called after the highly conserved residues in this region. The H box contains the conserved histidine (H) residue, which is the site for autophosphorylation. The receiver domain contains a conserved aspartate (D) residue. This D is the target for phosphorylation by FixL. The C-terminus of FixJ consists of a transcriptional activator (output) domain. Full-length (FixLc) and truncated (FixL3, 7 and 8) FixL proteins are fused at their N-terminus with glutathione S-transferase (GST) in order to facilitate their purification. The connecting lines between the S. meliloti and R. leguminosarum FixL proteins indicate the topological differences between the two HK domains.

 
In order to investigate the role of the C-terminal receiver domain, a fixL derivative was constructed with a TGA stop codon introduced between the HK domain and the receiver domain. Complementary oligonucleotides 9 and 10 (Table 2; Fig. 4a), encompassing the desired CAA to TGA mutation, were used in PCR in combination with two oligonucleotides hybridizing about 50 bp upstream from the NotI site in fixL and about 50 bp downstream from the kpnI site in the azu gene (Table 2; Fig. 4a). A 1797 bp DNA fragment was amplified from pTP95 DNA and cloned into pBluescript KS. This resulted in plasmid pAE8. The nucleotide sequence was determined. No changes, except the desired CAA to TGA mutation, were observed. The NotI–KpnI fragment from pAE8 was used to replace the original fixL fragment in pJSG103, after removal of the NotI site in the multiple cloning site (MCS). An additional frameshift was introduced downstream from the introduced stop codon by means of a fill-in religation reaction at the BstEII restriction site. A sequence-verified construct, pMKG8, was selected for further experiments.

Protein purification and in vitro phosphorylation.
The Bulk Glutathione S-Transferase (GST) Purification Module (Amersham Pharmacia Biotech) was used to purify GST–FixL fusion proteins from E. coli DH5{alpha} sonic lysates. Purification was done according to the manufacturer's instructions. The proteins were eluted from the resin with 10 mM reduced glutathione, 50 mM Tris/HCl, pH 8, and dialysed against 100 mM KCl, 0·1 mM EDTA, 0·1 mM DTT, 50 mM Tris/HCl, pH 8, and 50 % (v/v) glycerol. The proteins were stored at –20 °C until needed. Phosphorylation reactions were carried out as described by Tuckerman et al. (2001). No particular measures were taken to create anaerobic conditions for the phosphorylation experiments. The levels of phospho-FixL in the dried acrylamide gels were quantified with a Fujifilm FLA-3000 phosphoimager (Fuji Photo Film Co.).

Construction of the fixKL deletion mutant.
A VF39 derivative with a chromosomal deletion encompassing the fixK, fixL and azu genes was constructed. Plasmid pTP95 contains a 6936 bp EcoRI–PstI DNA fragment spanning the fixKL–NOQ' region from R. leguminosarum VF39. A 2505 bp HinDIII–KpnI fragment was deleted from pTP95. This deletion removed the entire coding region of the fixL gene, part of the C-terminus of fixK and part of the N-terminus of the azu gene. The remaining 4431 bp fragment was excised with BamHI and EcoRI and cloned into the chloramphenicol resistance gene of pAS269. The sacRB genes from pUM24 were introduced into the PstI site. This resulted in pMK97·9. Plasmid pMK97·9 was conjugated from E. coli S17-1 into R. leguminosarum VF39-TP4 (fixL : : GmR) (Patschkowski et al., 1996). Selecting for loss of gentamicin resistance resulted in a strain from which fixL and parts of the fixK and azu genes were deleted. This strain was designated VF39{Delta}KL (Fig. 4).

Complementation of VF39{Delta}KL.
The 1253 bp SalI–HinDIII fragment from pTP95 was cloned into pJP2 restricted with XhoI and HinDIII. This resulted in pMKJ-N, bearing a fixNc : : uidA gene fusion. A single NotI restriction site in this plasmid was removed by fill-in religation. The 2385 bp KpnI to HinDIII fragment from pTP95 containing the entire fixL gene was cloned into pMKJ-N. The resulting plasmid pMKJ38 contained a complete fixL gene and a fixK gene from which a 120 bp HinDIII fragment was missing. Another derivative, pMKJ7, was found to contain a complete fixK gene. This was probably due to a partial HinDIII digest. The KpnI to NotI fragment from pMKG8 was used to replace the 1681 bp KpnI–NotI fragment of pMKJ38 and pMKJ7. This resulted in pMKJ8 and pMKJ12, respectively.

These pMKJ plasmids could be used to complement the VF39{Delta}KL strain in trans and study the effect on the fixNc promoter. However, in order to avoid copy-number effects on gene regulation and be able to test other transcriptional gene fusions, we preferred chromosomal integration of the complementing fragments. Therefore, the KpnI to XbaI (which flank the SalI/XhoI hybrid site) fragments were excised from these pMKJ plasmids and transferred into pK18mob. The pK18mob derivatives (pMKK38-7, -8 and -12) were integrated into the VF39{Delta}KL genome by homologous recombination. The single crossover (see Fig. 4) should take place in the region of homology between the SalI site in fixNc and the HinDIII site in fixK. Successful integrations were verified by PCR with the fixNc and fixK2 oligonucleotide primers. Integration of plasmid pMKK7 into the VF39{Delta}KL genome resulted in strain VF39{Delta}KL-7, which has a wild-type genotype. Integration of pMKK38 resulted in VF39{Delta}KL-38, which has a fixK fixL+ genotype. Plasmid pMKK8 and pMKK12 both carry the truncated fixL gene (fixL{Delta}C). The resulting genotype of VF39{Delta}KL-8, therefore, is fixK fixL{Delta}C, and VF39{Delta}KL-12 is fixK+ fixL{Delta}C.

Culture conditions.
Aerobic conditions were achieved by shaking 100 ml cultures at 200 r.p.m. at 28 °C in 250 ml Duran bottles (Schott). Microaerobic induction of the fnrN promoter was achieved as follows. The cultures were grown aerobically to an OD600 of about 0·4. The undiluted aerobically grown cultures were transferred to 100 ml Duran bottles (100 ml cultures in a total volume of about 136 ml). The bottles were sealed and incubation was continued while shaking at 200 r.p.m. Samples were taken with a syringe through a rubber septum, maintaining the microaerobic conditions.

GUS activity was assayed as described by Prell et al. (2002). GUS activity is expressed as nmol p-nitrophenol (PNP) released per minute and per OD600 (nmol(minxOD600)–1).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Domain structure of the FixL protein
A FixL protein consists of at least two distinct domains (Fig. 1). The histidine kinase (HK) domain is conserved in all sensor proteins belonging to the family of two-component regulatory systems (Ronson et al., 1987). The H box is located at the N-terminal extreme of the HK domain and contains a conserved histidine residue, which is the target for autophosphorylation. Downstream of the H box is the nucleotide-binding region, which consists of an N, a G1, an F and a G2 box named according to the conserved amino acids in this domain (Parkinson and Kofoid, 1992). Typical for a FixL protein is the haem-containing PAS domain, which is involved in oxygen sensing and the regulation of the kinase activity of the downstream HK domain (Lois et al., 1993; Gilles-Gonzalez and Gonzalez, 1993). In order to obtain some indication of how widespread FixL genes are, we used the default BLASTP (Altschul et al., 1997) to search the NCBI GenBank database (www.ncbi.nlm.nih.gov) with the R. leguminosarum VF39 FixL protein sequence (GenBank Entrez protein ID: CAA94319). The low-complexity filter was suppressed and the search was limited to 250 alignments (March 2004). From all proteins that showed a continuous similarity with both the haem-containing PAS domain and the downstream HK domain, 13 proteins were selected that conserved the histidine residue at position 200 and the arginine residue at position 220 (positions according to the Bradyrhizobium japonicum FixL protein; Gong et al., 1998). Arginine 220 is thought to be involved in oxygen binding and is conserved in all known FixL proteins (Dunham et al., 2003). The isoleucines 215 and 216, which play an important role in the signal transduction between the haem and kinase domains (Mukai et al., 2000), were also highly conserved within these 13 proteins. These proteins encompassed all known Rhizobium FixL proteins: Rhizobium etli CFN42 AAG00949; S. meliloti NP 435916 and the FixL-related protein B95339; B. japonicum NP 769400; Azorhizobium caulinodans P26489 and Mesorhizobium loti NP 107078. The R. etli (previously R. leguminosarum bv. phaseoli) CNPAF512 FixL protein AAA96818 does not feature in this list. The PAS domain in this FixL protein differs from other FixL proteins and does not bind haem (D'hooghe et al., 1998). Putative FixL proteins were also identified in other Alphaproteobacteria’, such as Rhodopseudomonas palustris NP_949583, Caulobacter crescentus NP_419576 and Novosphingobium aromaticivorans ZP_00093963 and ZP_00095226. Other bacteria with putative FixL proteins included another proteobacterium (Magnetococcus sp. ZP_00044579), a Planctomycete (Pirellula sp. NP_866296) and a Bacteriodete (Cytophaga hutchinsonii ZP_00116841). An alignment of the PAS domains of these 13 proteins and the R. leguminosarum VF39 FixL showed high conservation of virtually all residues that are believed to be involved in ligand binding (Gong et al., 1998; Perutz et al., 1999; Mukai et al., 2000; Dunham et al., 2003; data not shown). Four out of these 14 proteins also contained a covalently linked C-terminal receiver domain. These are the R. leguminosarum and the R. etli CFN42 FixL proteins, the S. meliloti FixL-related protein and the putative FixL protein from Magnetococcus sp.

We observed that the R. leguminosarum and the R. etli CFN42 FixL proteins and the S. meliloti FixL-related protein have an unorthodox H box sequence: HDFNNLL. A typical type-1 HK domain contains the consensus motif HEhRTPh (h=conserved hydrophobic residue; Kim and Forst, 2001). The majority of the Rhizobium FixL proteins have a HELNQPL H box sequence. In particular, the first residue after the conserved histidine, which is usually a glutamate, is replaced by an aspartate, and the fifth residue, which is almost invariably a proline, is replaced by leucine. Furthermore, there are some topological differences, indicated in Fig. 1, in the spacing between the haem domain and the H box and between the H box and the downstream nucleotide-binding site. The spacing between the N and G1 boxes in the nucleotide-binding region is increased by 10–14 amino acids. All proteins from this BLAST search with similar unorthodox HK domains also possessed a covalently linked downstream receiver domain.

FixL is a haem-containing protein with autophosphorylation activity
R. leguminosarum FixL consists of at least four distinct domains (Fig. 1). Unlike the S. meliloti FixL, the R. leguminosarum protein does not feature significant transmembrane segments in its N-terminus. The protein is therefore probably cytoplasmic, rather than membrane-located. A SMART (Schultz et al., 2000) analysis shows that the R. leguminosarum FixL features another PAS domain at its N-terminus instead (Fig. 1). A number of truncated derivatives of FixL, lacking one or more of these domains, was constructed. Each FixL protein was fused with GST in order to facilitate its purification. The FixL proteins were overexpressed in E. coli and batch-purified using a glutathione–Sepharose matrix. FixLc is the complete FixL protein with GST fused to its N-terminus. A spectrophotometric scan of the purified FixLc protein (600–250 nm) featured an adsorption peak at 395 nm (data not shown). This was expected for a haem-containing protein, and indicates that the iron atom in the haem moiety is in the Fe3+ (ferric) form (Gilles-Gonzalez et al., 1994). Incubating the protein in the presence of 10 mM DTT under a N2 atmosphere resulted in a shift of this absorption peak to 431 nm. This wavelength indicates that the haem iron is in the reduced Fe2+ (ferrous) form. Both forms of the GST : : FixLc protein were tested for in vitro autokinase activity in the presence of [{gamma}-32P]ATP (Fig. 2a). The ferric form of the protein had significant autophosphorylation activity. Maximum levels of 32P incorporation were obtained in about 15 min, at which time about 6 % of the FixLc protein was phosphorylated. This level of phosphorylation of FixL is comparable to the level of phosphorylation of the S. meliloti FixL when incubated in absence of FixJ (Tuckerman et al., 2001). The Fe2+ form of the protein, which was incubated simultaneously under identical conditions, was much less active. Therefore, all other GST : : FixL derivatives were tested in vitro for autokinase activity in their ferric form.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 2. In vitro phosphorylation of FixL. (a) Phosphorylation of FixLc. The native FixLc protein, as purified, has an ODmax at 395 nm (FixLc395). Upon incubation with DTT in an anoxic environment, the absorption peak shifts to 432 nm (FixLc432). Both proteins were tested simultaneously for autophosphorylation. Samples were taken at 2, 4, 8, 16 and 32 min after addition of radiolabelled ATP. The products were separated on a 12 % polyacrylamide gel and visualized through staining with Coomassie Brilliant Blue. The gel was dried and exposed overnight to a phosphoimage screen. The screen was scanned in a phosphoimager to visualize the distribution of radioactivity in the gel. The relative intensities of the radioactive spots are plotted against reaction time. (b) Phosphotransfer to the receiver domain. Three time-course reactions, similar to the one described in (a), were performed simultaneously. FixL3 and FixL8 were tested in the ferric form. The relative intensity of the radioactive spots is plotted against the reaction time. Samples were taken at 2, 4, 8, 16, 32 and 64 min after addition of the radiolabelled ATP. Reaction 1 (1) contained FixL3 ({square}) and FixL7 ({blacksquare}). FixL3 autophosphorylates efficiently and phosphorylates FixL7 in trans. Reaction 2 (2) contained FixL8 ({triangleup}) and FixL7 ({blacktriangleup}). FixL8 does not autophosphorylate, but efficiently phosphorylates FixL7 in trans. Reaction 3 (3) contained the FixL8 protein only ({circ}). Autophosphorylation of FixL8 can be observed in the absence of FixL7.

 
The C-terminal receiver domain is phosphorylated by the HK domain
The N-terminal PAS domain of FixL was removed by fusing the GST moiety in FixL3 directly to the haem domain.

The truncated FixL3 protein still had significant autokinase activity (Fig. 3, lanes 1 and 2). This indicated that the N-terminal PAS domain is not essential for in vitro autophosphorylation activity. However, the incorporation of 32P was significantly lower than that of the FixLc protein (not shown).



View larger version (57K):
[in this window]
[in a new window]
 
Fig. 3. Phosphotransfer to the receiver domain. FixL7 and FixL3 were tested for autophosphorylation and intermolecular phosphotransfer. Reactions were carried out as described in Fig. 2. (a) Dried gel stained with Coomassie Brilliant Blue. Lanes 1 and 2, reactions containing the FixL3 protein (6 µg protein, reaction time 10 min); lane 3, reaction containing FixL7 (6 µg protein, reaction time 10 min); lanes 4–9, time-course of a large-scale reaction containing FixL3 (12 µg protein per sample) and FixL7 (6 µg protein per sample). Samples were taken at 0·5, 1, 2, 4, 8 and 16 min after addition of radiolabelled ATP. (b) Phosphoimage. Lanes 1 and 2, FixL3 autophosphorylates efficiently; lane 3, FixL7 does not autophosphorylate; lanes 4–9, FixL3 efficiently phosphorylates FixL7 in trans. S, protein size standard (Multimark, NOVEX). Protein sizes are indicated to the right of panel (a).

 
A gene fusion of GST to the receiver domain resulted in FixL7. FixL7 had no autokinase activity (lane 3). Since this protein lacks the HK domain, the lack of autokinase activity was expected. However, the receiver domain was efficiently phosphorylated in trans in the presence of FixL3 (Fig. 3, lanes 4–9) or FixLc (not shown).

FixL8 is a derivative of FixL3 which contains a TGA stop codon between the HK- and the C-terminal receiver domains. As a result, FixL8 lacks both the N-terminal PAS domain and the C-terminal receiver domain. FixL8 had low autokinase activity, despite the fact that the HK domain was not altered in this protein (Fig. 2b). When FixL8 and FixL7 were incubated together, FixL7 became efficiently phosphorylated, whereas virtually no radioactivity was incorporated into FixL8. This indicates that the HK domain of FixL8 is functional and that the phosphoryl signal is transferred rapidly to the receiver domain in FixL7.

FixL is required for microaerobic induction of the fnrN promoter
A transcriptional gene fusion of the fnrN promoter with the uidA (GUS) reporter gene was constructed in pJP2. This plasmid (pJP-R) was used to monitor the expression of the fnrN promoter under various physiological conditions and in different genetic backgrounds. The pJP-R reporter was introduced into the wild-type (VF39) and the fixKL mutant strain (VF39{Delta}KL). No significant activity of the fnrN promoter was obtained under aerobic conditions, regardless of the strain used (t=0, Fig. 5). In a standard sealed-bottle experiment, the headspace was about 36 ml. This amount of air allowed the cultures to grow for approximately 6 h until anaerobic conditions were reached. No differences in growth rate were observed between the various strains. In the wild-type background, the fnrN promoter was induced after about 1 h, and GUS activity rapidly increased to a maximum level of about 250 units (Fig. 5). No induction of the fnrN promoter was observed in the fixKL mutant.



View larger version (20K):
[in this window]
[in a new window]
 
Fig. 5. Microaerobic induction of the fnrN promoter. All strains contained the pJP-R plasmid bearing a transcriptional gene fusion of the fnrN promoter to the {beta}-glucuronidase (GUS) gene. WT, R. leguminosarum bv. viciae VF39 (wild-type); L, VF39{Delta}KL (fixKL deletion strain); L38, VF39{Delta}KL-38 (fixL complemented); L8, VF39{Delta}KL-8 (fixL{Delta}C). The unmarked solid line represents the growth of the wild-type strain in this experiment. No differences in growth rates could be observed between these strains. OD600 is indicated on the left hand y axis. GUS activity, expressed in nmol p-nitrophenol produced per min and per OD600 unit, is indicated on the right hand y axis. Each strain was tested for microaerobic induction of the fnrN promoter at least three times. The curves shown in this figure are all derived from one experiment.

 
The receiver domain is important for fnrN expression
In order to establish if the induction of the fnrN promoter requires the presence of one or both of the fixK and fixL genes, we set out to complement the mutation in strain VF39{Delta}KL. Complementation was achieved by integration of pK18mob derivatives containing various portions of the fixKL operon into the VF39{Delta}KL genome (Fig. 4). When VF39{Delta}KL was complemented with pMKK7 (fixK+, fixL+), expression of the fnrN promoter was restored to wild-type levels (data not shown). The same was true for pMKK38 (fixK, fixL+; Fig. 5). This demonstrated that a functional copy of the R. leguminosarum fixL gene, but not the fixK gene, is required for microaerobic induction of the fnrN promoter.

It is possible that the phosphorelay pathway leads from the histidine in FixL, via a receiver domain in another transcriptional regulator protein, to the fnrN promoter. In that case, the C-terminal receiver domain of FixL may be dispensable. To investigate this possibility, VF39{Delta}KL was complemented with the truncated fixL{Delta}C gene. The presence of pMKK8 (fixK, fixL{Delta}C; Fig. 5) or pMKK12 (fixK+, fixL{Delta}C; not shown) failed to restore fnrN expression. These results indicate that the C-terminal receiver domain of the FixL protein does play an important role in the regulation of the fnrN promoter.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Most micro-organisms, even obligate aerobes such as R. leguminosarum, must be able to cope with conditions of limited oxygen availability. Many different systems have evolved to adapt gene expression to oxygen availability (Bauer et al., 1999). Our search for haem-containing oxygen-binding sensor (FixL-like) proteins identified 14 candidates. Seven of these were proteins from rhizobial species. The role and function of these FixL proteins is well documented. The other FixL-like proteins have emerged from genome sequencing projects and nothing is known about the actual role and function of these proteins. This is, for example, also the case for the S. meliloti gene encoding a FixL-related protein. This protein is a hybrid protein, like the R. leguminosarum VF39 FixL. In this context, it is notable that in S. meliloti several oxygen-regulated genes have been identified that are not under the control of the FixLJ system (Trzebiatowski et al., 2001). None of these genes is essential for symbiotic N2 fixation, suggesting that they may play a role under free-living microaerobic conditions.

The R. leguminosarum bv. viciae VF39 FixL protein was tested for in vitro autophosphorylation activity. The purified full-length FixLc protein featured an absorption peak at 395 nm. This indicated that the iron in the haem moiety was in the ferric form. In this form, the protein had significant autokinase and phosphorylation activity. Reducing the protein to the ferrous form by incubating with DTT under a N2 atmosphere had a negative effect on the autokinase activity. At first glance, this contradicts earlier reports that S. meliloti FixL activity is correlated to the electronic spin state of the haem iron (Gilles-Gonzalez et al., 1995). Recently, it has been reported that the Cys301 residue (located 16 amino acids downstream from the conserved histidine in the HK domain) in S. meliloti FixL causes an aberrant inactivation of the ferric form of the protein. A C301A mutation of the protein results in a FixL that is equally active for autokinase and FixJ phosphorylation in the ferric form as well as in the unliganded ferrous form (Akimoto et al., 2003). The Cys301 residue is not conserved in the R. leguminosarum FixL, nor is it conserved in any of the other FixL-like proteins. Therefore, it is not surprising that FixL is fully active in the ferric form. It is possible that, under our experimental conditions, a significant part of the reduced protein was liganded with O2. Oxygenation of the ligand may account for the observed reduction in autokinase activity of the ferrous form of the protein.

The function of the N-terminal domain in the R. leguminosarum VF39 FixL protein is not known. PAS domains occur in many sensory proteins and are generally involved in signal sensing by means of an associated cofactor (Taylor and Zhulin, 1999). The N-terminal PAS domain is not essential for autokinase activity. This, however, does not exclude a possible regulatory role for FixL activity. Comparison of phosphorylation rates is difficult, due to the short half-life of the 32P isotope. Conditions of purification and storage may also affect the activity of the FixL proteins. The C-terminal receiver domain had no autokinase activity, as expected. The receiver could be phosphorylated efficiently in trans by FixL3 and the full-length protein. This indicates that, in these proteins, intramolecular transfer of the phosphoryl from the histidine residue in the HK domain to the aspartate residue in the receiver domain (i.e. turnover) may take place. Deleting the C-terminal receiver domain from FixL3 resulted in a protein (FixL8) with low autokinase activity. Nevertheless, the capacity to phosphorylate the receiver domain in trans was retained. In the presence of FixL7, virtually all radioactivity ended up in the receiver domain, and no phosphorylation of the FixL8 protein could be observed. This suggests that the histidine residue in the kinase domain is only phosphorylated transiently, and that predominantly the aspartate residue in the receiver domain is phosphorylated. In the case of the S. meliloti FixL and FixJ proteins, it has been observed that the autokinase and transfer reactions are much faster and more efficient when the two proteins are allowed to form a complex before the ATP is added (Tuckerman et al., 2001, 2003). In hybrid proteins, both domains are part of the same protein and therefore very likely always to form a complex.

We observed that the R. leguminosarum FixL has an unorthodox H box sequence and a topologically altered nucleotide-binding region. All HK proteins with a similar unorthodox HK domain that emerged from the BLASTP search also had a covalently linked receiver module immediately downstream of the HK domain. This suggests that there is a structural difference between the phosphorelay mechanism in these hybrid proteins and the phosphorelay in two-component systems in which the HK and receiver domains are located on separate proteins. This difference may be reflected in the low level of autophosphorylation of the R. leguminosarum FixL{Delta}C protein. In hybrid proteins, there is always a receiver domain present to accept the phosphoryl signal. In two-component systems, a receiver domain may not always be in the vicinity, and the phosphorylated histidine may have to be stabilized by surrounding residues.

The R. leguminosarum bv. viciae VF39 FixL protein plays an important role in the regulation of the fnrN promoter. The fnrN promoter is easily induced under microaerobic conditions, and vigorous shaking in unsealed flasks was required to repress the promoter. When oxygen becomes limited, the fnrN : : uidA gene fusion is induced. A significant reduction in promoter activity was observed in the {Delta}fixKL strain. Wild-type levels of induction were restored when this strain was complemented with a full-length copy of the fixL gene. Complementation of the fixK mutation was not required. This suggests that FixK is not involved in the positive regulation of the fnrN promoter. When the C-terminal receiver domain was deleted from the fixL gene, induction of the fnrN promoter could not be restored. This indicates that the C-terminal receiver domain of FixL is essential for regulation of the fnrN promoter under free-living microaerobic conditions.

It is clear that the FixL C-terminal receiver domain is required for fast induction of the fnrN promoter when oxygen levels become limited. It is, however, not clear how FixL exerts this effect on the fnrN promoter. From its domain structure, it is unlikely that FixL itself interacts with the fnrN promoter. It has been shown that fnrN is positively autoregulated and requires the alternative transcription factor RpoN (Clark et al., 2001). In another strain of R. leguminosarum bv. viciae, UPM791, FnrN positively and negatively regulates its own transcription (Colombo et al., 2000). In our hands, fnrN shows little or no autoregulation. Furthermore, FixK could also interact with the anaeroboxes in the fnrN promoter region. Although, in this study, FixK was not essential for fnrN activity, it is still possible that this regulator negatively influences fnrN transcription. FixL may, directly or indirectly, connect with any of these transcription factors in order to exert its effect on the fnrN promoter. Furthermore, under different physiological conditions, other regulatory pathways may operate, and additional environmental signals may have an effect on proteins involved in the oxygen regulation of gene expression.


   ACKNOWLEDGEMENTS
 
We thank Maria Krämer and Ayse Alcali for technical assistance, Jörg Schumacher for initializing the protein work and Andreas Schlüter and Alexandra Ehrlach for plasmid construction.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Akimoto, S., Tanaka, A., Nakamura, K., Shiro, Y. & Nakamura, H. (2003). O2-specific regulation of the ferrous haem-based sensor kinase FixL from Sinorhizobium meliloti and its aberrant inactivation in the ferric form. Biochem Biophys Res Commun 304, 136–142.[CrossRef][Medline]

Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–3402.[Abstract/Free Full Text]

Bauer, C. E., Elsen, S. & Bird, T. H. (1999). Mechanisms for redox control of gene expression. Annu Rev Microbiol 53, 495–523.[CrossRef][Medline]

Beringer, J. E. (1974). R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84, 188–198.[Medline]

Clark, S. R., Oresnik, I. J. & Hynes, M. F. (2001). RpoN of Rhizobium leguminosarum bv. viciae strain VF39SM plays a central role in FnrN-dependent microaerobic regulation of genes involved in nitrogen fixation. Mol Gen Genet 264, 623–633.[CrossRef][Medline]

Colombo, M. V., Gutierrez, D., Palacios, J. M., Imperial, J. & Ruiz-Argueso, T. (2000). A novel autoregulation mechanism of fnrN expression in Rhizobium leguminosarum bv viciae. Mol Microbiol 36, 477–486.[CrossRef][Medline]

Colonna-Romano, S., Arnold, W., Schlüter, A., Boistard, P., Pühler, A. & Priefer, U. B. (1990). An Fnr-like protein encoded in Rhizobium leguminosarum biovar viciae shows structural and functional homology to Rhizobium meliloti FixK. Mol Gen Genet 223, 138–147.[Medline]

David, M., Daveran, M. L., Batut, J., Dedieu, A., Domergue, O., Ghai, J., Hertig, C., Boistard, P. & Kahn, D. (1988). Cascade regulation of nif gene expression in Rhizobium meliloti. Cell 54, 671–683.[Medline]

D'hooghe, I., Michiels, J. & Vanderleyden, J. (1998). The Rhizobium etli FixL protein differs in structure from other known FixL proteins. Mol Gen Genet 257, 576–580.[CrossRef][Medline]

Dunham, C. M., Dioum, E. M., Tuckerman, J. R., Gonzalez, G., Scott, W. G. & Gilles-Gonzalez, M. A. (2003). A distal arginine in oxygen-sensing haem-PAS domains is essential to ligand binding, signal transduction, and structure. Biochemistry 42, 7701–7708.[CrossRef][Medline]

Fischer, H. M. (1994). Genetic regulation of nitrogen fixation in Rhizobia. Microbiol Rev 58, 352–386.[Medline]

Gilles-Gonzalez, M. A. & Gonzalez, G. (1993). Regulation of the kinase activity of haem protein FixL from the two-component system FixL/FixJ of Rhizobium meliloti. J Biol Chem 268, 16293–16297.[Abstract/Free Full Text]

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, 170–172.[CrossRef][Medline]

Gilles-Gonzalez, M. A., Gonzalez, G., Perutz, M. F., Kiger, L., Marden, M. C. & Poyart, C. (1994). Haem-based sensors, exemplified by the kinase FixL, are a new class of haem protein with distinctive ligand binding and autoxidation. Biochemistry 33, 8067–8073.[Medline]

Gilles-Gonzalez, M. A., Gonzalez, G. & Perutz, M. F. (1995). Kinase activity of oxygen sensor FixL depends on the spin state of its haem iron. Biochemistry 34, 232–236.[Medline]

Gong, W., Hao, B., Mansy, S. S., Gonzalez, G., Gilles-Gonzalez, M. A. & Chan, M. K. (1998). Structure of a biological oxygen sensor: a new mechanism for haem-driven signal transduction. Proc Natl Acad Sci U S A 95, 15177–15182.[Abstract/Free Full Text]

Green, J., Scott, C. & Guest, J. R. (2001). Functional versatility in the CRP-FNR superfamily of transcription factors: FNR and FLP. Adv Microb Physiol 44, 1–34.[Medline]

Hanahan, D. (1985). In DNA Cloning, Vol. 1, pp. 109–135. Edited by D. M. Glover. Oxford: IRL Press.

Kim, D. & Forst, S. (2001). Genomic analysis of the histidine kinase family in bacteria and archaea. Microbiology 147, 1197–1212.[Medline]

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, 4370–4375.[Abstract/Free Full Text]

Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Monson, E. K., Weinstein, M., Ditta, G. S. & Helinski, D. R. (1992). The FixL protein of Rhizobium meliloti can be separated into a haem-binding oxygen-sensing domain and a functional C-terminal kinase domain. Proc Natl Acad Sci U S A 89, 4280–4284.[Abstract]

Mukai, M., Nakamura, K., Nakamura, H., Iizuka, T. & Shiro, Y. (2000). Roles of Ile209 and Ile210 on the haem pocket structure and regulation of histidine kinase activity of oxygen sensor FixL from Rhizobium meliloti. Biochemistry 39, 13810–13816.[CrossRef][Medline]

Parkinson, J. S. & Kofoid, E. C. (1992). Communication modules in bacterial signaling proteins. Annu Rev Genet 26, 71–112.[CrossRef][Medline]

Patschkowski, T., Schlüter, A. & Priefer, U. B. (1996). Rhizobium leguminosarum bv. viciae contains a second fnr/fixK-like gene and an unusual fixL homologue. Mol Microbiol 21, 267–280.[CrossRef][Medline]

Perutz, M. F., Paoli, M. & Lesk, A. M. (1999). FixL, a haemoglobin that acts as an oxygen sensor: signalling mechanism and structural basis of its homology with PAS domains. Chem Biol 6, 291–297.[CrossRef]

Preisig, O., Anthamatten, D. & Hennecke, H. (1993). Genes for a microaerobically induced oxidase complex in Bradyrhizobium japonicum are essential for a nitrogen-fixing endosymbiosis. Proc Natl Acad Sci U S A 90, 3309–3313.[Abstract]

Preisig, O., Zufferey, R., Thony-Meyer, L., Appleby, C. A. & Hennecke, H. (1996). A high-affinity cbb3-type cytochrome oxidase terminates the symbiosis-specific respiratory chain of Bradyrhizobium japonicum. J Bacteriol 178, 1532–1538.[Abstract]

Prell, J., Boesten, B., Poole, P. & Priefer, U. B. (2002). The Rhizobium leguminosarum bv viciae VF39 {gamma}-aminobutyrate (GABA) aminotransferase gene (gabT) is induced by GABA and highly expressed in bacteroids. Microbiology 148, 615–623.[Medline]

Priefer, U. B. (1989). Genes involved in lipopolysaccharide production and symbiosis are clustered on the chromosome of Rhizobium leguminosarum biovar viciae VF39. J Bacteriol 171, 6161–6168.[Medline]

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, 6867–6872.[Abstract]

Ried, J. L. & Collmer, A. H. (1987). An nptI-sacB-sacR cartridge for constructing directed, unmarked mutations in Gram-negative bacteria by marker exchange-eviction mutagenesis. Gene 57, 239–246.[CrossRef][Medline]

Ronson, C. W., Nixon, B. T. & Ausubel, F. M. (1987). Conserved domains in bacterial regulatory proteins that respond to environmental stimuli. Cell 49, 579–581.[CrossRef][Medline]

Schäfer, A., Tauch, A., Jager, W., Kalinowski, J., Thierbach, G. & Pühler, A. (1994). Small mobilizable multi-purpose cloning vectors derived from the Escherichia coli plasmids pK18 and pK19: selection of defined deletions in the chromosome of Corynebacterium glutamicum. Gene 145, 69–73.[CrossRef][Medline]

Schlüter, A., Patschkowski, T., Quandt, J., Selinger, L. B., Weidner, S., Kramer, M., Zhou, L., Hynes, M. F. & Priefer, U. B. (1997). Functional and regulatory analysis of the two copies of the fixNOQP operon of Rhizobium leguminosarum strain VF39. Mol Plant–Microbe Interact 10, 605–616.[Medline]

Schultz, J., Copley, R. R., Doerks, T., Ponting, C. P. & Bork, P. (2000). SMART: a web-based tool for the study of genetically mobile domains. Nucleic Acids Res 28, 231–234.[Abstract/Free Full Text]

Simon, R., Priefer, U. B. & Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria. Bio/Technology 1, 784–791.

Taylor, B. L. & Zhulin, I. B. (1999). PAS domains: internal sensors of oxygen, redox potential, and light. Microbiol Mol Biol Rev 63, 479–506.[Abstract/Free Full Text]

Trzebiatowski, J. R., Ragatz, D. M. & de Bruijn, F. J. (2001). Isolation and regulation of Sinorhizobium meliloti 1021 loci induced by oxygen limitation. Appl Environ Microbiol 67, 3728–3731.[Abstract/Free Full Text]

Tuckerman, J. R., Gonzalez, G. & Gilles-Gonzalez, M. A. (2001). Complexation precedes phosphorylation for two-component regulatory system FixL/FixJ of Sinorhizobium meliloti. J Mol Biol 308, 449–455.[CrossRef][Medline]

Tuckerman, J. R., Gonzalez, G., Dioum, E. M. & Gilles-Gonzalez, M. A. (2003). Ligand and oxidation-state specific regulation of the haem-based oxygen sensor FixL from Sinorhizobium meliloti. Biochemistry 41, 6170–6177.[CrossRef]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103–119. Erratum in Gene 114, 81–83.

Zhulin, I. B., Taylor, B. L. & Dixon, R. (1997). PAS domain S-boxes in Archaea, Bacteria and sensors for oxygen and redox. Trends Biochem Sci 22, 331–333.[CrossRef][Medline]

Received 12 May 2004; revised 22 July 2004; accepted 4 August 2004.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Boesten, B.
Articles by Priefer, U. B.
Articles citing this Article
PubMed
PubMed Citation
Articles by Boesten, B.
Articles by Priefer, U. B.
Agricola
Articles by Boesten, B.
Articles by Priefer, U. B.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.