Identification of a NifL-like protein in a diazotroph of the ß-subgroup of the Proteobacteria, Azoarcus sp. strain BH72c

Tanja Egenera,1, Abhijit Sarkar2, Dietmar E. Martinb,2 and Barbara Reinhold-Hurek1,2

Max-Planck-Institute for Terrestrial Microbiology, Group Symbiosis Research, Karl-von-Frisch-Strasse,D-35043 Marburg, Germany1
University of Bremen, Faculty of Biology and Chemistry, Laboratory of General Microbiology, Postfach 330440, D-28334 Bremen, Germany2

Author for correspondence: Barbara Reinhold-Hurek. Tel: +49 421 218 2370. Fax: +49 421 218 4042. e-mail: breinhold{at}uni-bremen.de


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
NifA, the transcriptional activator of nitrogenase (nif) genes, has up to now been described to be regulated in its activity via the sensor NifL only for members of the {gamma}-subgroup of the Proteobacteria. This paper reports a functionally similar NifL-like protein outside this group in Azoarcus sp. strain BH72, a diazotrophic grass endophyte belonging to the ß-subgroup of the Proteobacteria. Its structural genes for nitrogenase (nifHDK) are regulated in response to combined nitrogen and O2 and expressed endophytically inside rice roots. In order to characterize nitrogen-regulatory genes, an Azoarcus sp. BH72 genomic library was used to select cosmids that complemented a nifA mutation in Azotobacter vinelandii. Sequence analysis of the 3·4 kb genomic region complementing nifA showed two ORFs with sequence identities of 44% to NifL and 61% to NifA of Azotobacter vinelandii. According to Northern blot and reverse transcriptase PCR analysis, the nifLA transcript was more abundant at low combined nitrogen and O2 levels, results which were corroborated by GUS (ß-glucuronidase) assays using a transcriptional nifL::gusA fusion. N2 fixation was abolished in a NifLA- and a NifA- mutant, wild-type fixation being restored by nifLA in trans. The NifLA- mutant also failed to activate nifH::gus expression, indicating that NifA is the obligate transcriptional activator for nifHDK. A nifL mutant was diazotrophic and did not show repression of nifH::gusA by ammonium or O2, suggesting that NifL of Azoarcus sp. strain BH72 has a similar role in inactivating NifA in response to O2 and combined nitrogen as NifL in bacteria of the {gamma}-Proteobacteria.

Keywords: nitrogenase, gene expression, transcriptional activator, NifA, NifL

Abbreviations: GUS, ß-glucuronidase

c The GenBank accession number for the sequence determined in this work is AF518560.

a Present address: Albert-Ludwigs-University Freiburg, Plant Biotechnology, Sonnenstrasse 5, D-79104 Freiburg, Germany.

b Present address: Department of Biochemistry, Biozentrum University of Basel, CH-4056 Basel, Switzerland.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Biological N2 fixation is a highly energy-consuming process which is tightly controlled in free-living bacteria. Commonly, the structural genes of the nitrogenase complex, nifHDK, as well as other nif genes necessary for maturation and function of nitrogenase, are controlled by the {sigma}54-dependent transcriptional activator NifA in diazotrophic proteobacteria. Its activity is affected by the cellular nitrogen status and the O2 concentration; however, bacteria differ in the mechanisms mediating these responses.

In diazotrophs such as proteobacteria of the {alpha}-subgroup or Herbaspirillum seropedicae belonging to the ß-subgroup (Souza et al., 1991 ), the NifA proteins are apparently directly responsive to O2. These proteins show a conserved cysteine motif in the central domain, not present in NifA proteins of the {gamma}-Proteobacteria, which is probably the site of a redox-sensitive Fe–S cluster (Dixon, 1998 ; Fischer et al., 1988 ). Diazotrophs belonging to the {gamma}-Proteobacteria such as Azotobacter vinelandii and Klebsiella pneumoniae are characterized by the NifL/NifA two-component regulatory system, with NifL being the sensor inhibiting NifA activity in response to O2 (Dixon, 1998 ). Stoichiometric amounts of both proteins are needed to ensure proper transcriptional regulation (Dixon, 1998 ; Govantes et al., 1996 ). For O2 sensing, the flavoprotein NifL inhibits NifA activity in the oxidized form (Dixon, 1998 ; Hill et al., 1996 ).

The mechanism by which the cellular nitrogen status is sensed and the signal is transmitted is more complex and may vary considerably in different diazotrophs. One level of control is the transcriptional regulation of nifA itself, which may be nitrogen regulated via the two-component regulatory system NtrBC as in K. pneumoniae (Drummond et al., 1983 ) or H. seropedicae (Souza et al., 2000 ). At another level, the activity of NifA is modulated, PII-like proteins being central signal-transmitter proteins. Depending on the internal nitrogen status of the cell, a bifunctional uridylyl-transferase/hydrolase covalently modifies or demodifies the PII-like protein. Under conditions of nitrogen deficiency, the PII-like proteins in enteric bacteria occur mainly in the uridylylated form. They act as molecular switches depending on their state of modification (for a review, see Arcondéguy et al., 2001 ). Most proteobacteria harbour two paralogous gene copies of PII-like proteins (glnB/glnK) (Ninfa & Atkinson, 2000 ). In H. seropedicae and Azospirillum brasilense, GlnB is required to maintain the active form of NifA, either by direct interaction or by involvement of an as yet unkown protein (Arsène et al., 1996 , 1999 ; Souza et al., 1999 ). In {gamma}-Proteobacteria, NifA activity is mediated by NifL in response to combined nitrogen. Under nitrogen-limiting conditions, GlnK is required to relieve the inhibitory effect of NifL on NifA in K. pneumoniae (He et al., 1998 ; Jack et al., 1999 ). Interaction of PII-like proteins of Escherichia coli with Azotobacter vinelandii NifL was demonstrated in an in vitro system; however, E. coli GlnB (and Azotobacter vinelandii PII) rather than GlnK stimulated the inhibitory function of NifL in the non-uridylylated form (Little et al., 2000 ).

Azoarcus sp. strain BH72 is an endophyte of grasses (Hurek et al., 1994b ) which belongs to the ß-Proteobacteria (Reinhold-Hurek et al., 1993b ). N2 fixation and nifHDK transcription occur only under microaerobic and nitrogen-limiting conditions (Egener et al., 1999 ; Huwrek et al., 1987 ). This diazotroph shows the capacity of endophytic N2 fixation: the structural genes of nitrogenase nifHDK were found to be expressed and translated in the aerenchyma of rice seedlings (Egener et al., 1999 ). Under certain culture conditions including very low O2 concentrations, the cells can shift into a state of very high and efficient N2 fixation called hyperinduction (Hurek et al., 1994a ), forming novel intracytoplasmic membrane stacks (diazosomes) with which the iron protein of nitrogenase is associated (Hurek et al., 1995 ). Therefore, we are interested in unravelling the signal transduction cascade for nif gene regulation of strain BH72. We report here the occurrence of a functionally similar NifL-like protein outside the {gamma}-Proteobacteria.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacterial strains and plasmids.
The strains and plasmids used in this study are listed in Table 1. Additional subclones are shown in Fig. 1. Azotobacter vinelandii strains were kindly provided by C. Kennedy, University of Arizona, Tuscon.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 


View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Restriction map and subclones of pTEA24. The position of the ORFs are indicated. pTEA24 is a HindIII/EcoRI subclone in pBKSII of pTE2 derived from the cosmid library. Depicted below are subclones of 0·5–2kb inserts generated by PstI, XhoI and SalI and cloned into pBluescript SK. Sites for important and cloning-relevant sites are shown. A, ApaI; E, EcoRI; EV, EcoRV; HIII, HindIII; P, PstI; K, KpnI; S, SalI; X, XhoI.

 
Culture media and growth conditions.
Unless stated otherwise, Azoarcus sp. BH72 was grown at 37 °C in VM medium supplemented with ethanol (Reinhold-Hurek et al., 1993a , b ). N2-fixing cells were grown on N-free SM medium (Reinhold et al., 1985 ) under microaerobic conditions in Erlenmeyer flasks (Karg & Reinhold-Hurek, 1996 ). To study the effect of O2 on gene expression, cells precultured on SM medium with combined nitrogen (0·05% NH4Cl, 0·01% yeast extract) were washed in N-free medium and incubated under different O2 concentrations at an OD578 of 0·2 as described previously (Egener et al., 1999 ). To test the influence of ammonium on nitrogenase activity and nif gene expression, the following modifications were applied: cells were incubated aerobically at OD578 0·5 in N-free medium for 1 h followed by an incubation for 3 h under a headspace of 1% O2 and 10% acetylene in N2. Before cells were harvested to measure GUS activity (see below), the ethylene formed was measured gas chromatographically. For strain BHNLK the nitrogen-starvation phase was omitted but the incubation time was prolonged to 4 h.

Gas chromatography.
To determine O2 concentrations and ethylene formation, an HRGC-4000A (Konik, Barcelona, Spain) gas chromatograph was used as described previously (Egener et al., 1999 ).

Techniques for DNA and RNA manipulation.
DNA and RNA analysis was carried out according to standard procedures (Ausubel et al., 1987 ; Hurek et al., 1993 ; Reinhold-Hurek et al., 1993a ). Homologous DNA gene probes for Southern and Northern blot analysis were digoxigenin (DIG)-labelled in a PCR reaction using the DIG Labelling and Detection Kit (Boehringer Mannheim). The nifA probe was amplified with T3/T7 primers using a 0·5 kb PstI subclone (pP05) of pTEA24 as template. Primers for nifH were TH25/TH26 (Hurek et al., 1997b ), and for 16S rDNA TH3/TH5 (Hurek et al., 1993 ). RNA was isolated from exponentially growing cells, and Northern blot analysis carried out as previously described (Reinhold-Hurek et al., 1993a ). Reverse transcription PCR (RT-PCR) was carried out on RNA extracted by peqGOLD TriFast (peqLab) (Egener et al., 2001 ), with 1 µg RNA using primer nifArev3RT (TCGTCCAGGTGCTCGCGGCTG) and Ready-to-go-beads (AmershamPharmacia Biotech) for reverse transcription at 42 °C for 30 min. Amplification of nifA cDNA was carried out using primers nifAfor1RT (ATGAGCGCGGCCGGTCCGATG) and nifArev2RT (CACGGTTTCGTGCCCGGCGCG) for 18–28 cycles of 1 min 95 °C, 1 min 65 °C, 1 min 72 °C; samples were taken after different cycle numbers. For RT-PCR amplification of nifLA, primers RTnifLAfor (GAGAACGGCCAGGTCGACGTGGA, positions 1772–1792) and RTnifArev (GTTGAAGCCGCACTCCTCGTCGAGCA, positions 2194–2170) were used for 35 cycles of 95 °C for 1 min, 62 °C for 1 min and 72 °C for 1 min. Products were separated on 1·2% agarose gels. RT-PCR of 16S rRNA was carried out from 10 ng RNA with primers 1401rev (CGGTGTGTACAAGACCC) for reverse transcription and 104f (GGCGAACGGGTGMGTAAYGCACTGG) and 1346rev (TAGCGATTCCGACTTCA) for PCR amplification: 1 min 95 °C, 2 min 65 °C, 2 min 72 °C.

For primer extension analysis (Egener et al., 2001 ), 15 µg RNA extracted with the peqGOLD TriFast Kit (peqLab) was used as starting template, along with the Cy5 labelled primer (Cy5PEnifLArev: GATCGCCGACTGCTCCACGG). The reaction mix was denatured at 72 °C for 3 min followed by gradual cooling to 42 °C, then the dNTP mix and AMV reverse transcriptase were added and incubated for 30 min. The reaction mix was then extracted with phenol/chloroform followed by ethanol precipitation of the single-stranded cDNA. The product was dissolved in TE (10 mM Tris, 1 mM EDTA) and analysed on an automated sequencer (ALFexpress; AmershamPharmacia Biotech) in parallel with a sequencing reaction carried out with the same primer and plasmid template pS08.

DNA sequencing and computational analysis.
Plasmid DNA was sequenced with Cy5 labelled primers (T3 and T7) using an automated sequencer (ALFexpress, AmershamPharmacia Biotech) as described by Hurek et al. (1997a ). Sequences determined for both strands were aligned using the DNAstar software and compared to databases using BLAST (Altschul et al., 1990 ). Domain searches were carried out with the SMART (Schultz et al., 1998 ) and Pfam (Sonnhammer et al., 1997 ) programs. The sequence of nifLA was submitted to GenBank (accession no. AF518560).

Construction of a cosmid library of Azoarcus sp. BH72 and triparental mating.
Genomic DNA of Azoarcus sp. BH72 was partially digested with Sau3AI and cloned into the BamHI site of the cosmid vector pLAFR3 (Staskawicz et al., 1987 ). The ligation mix was packaged into lambda phages using the DNA Packaging Kit (Boehringer Mannheim) and transfected into E. coli DH5{alpha}. A total of 1440 colonies were picked and stored as glycerol stocks at -80 °C. Triparental mating in Azotobacter vinelandii was carried out according to Page & Sadoff (1976) using a helper plasmid [E. coli(pRK2013)] with a donor:helper:reipient ratio of 1:1:100. When the donor carried different inserts (i.e. a library) the ratio was altered to 5:1:100. The selection medium, containing tetracycline (12 µg ml-1), was N-free BS medium (Newton et al., 1953 ) for Azotobacter vinelandii UW1. For conjugative plasmid transfer into Azoarcus sp., recipients were grown in liquid VM medium, mixed in a ratio of 1:1:500 and spread as a thin liquid layer on KON agar plates (SM medium supplemented with 5 g yeast extract l-1 and 1 g NaCl l-1). After 5–7 h incubation at 37 °C, cells were scraped off and plated in serial dilutions on SEL medium (SM medium containing 6 ml ethanol l-1 instead of potassium malate and 1 g KNO3 l-1 as sole nitrogen source) with 12 µg tetracycline ml-1.

Construction of Azoarcus sp. BH72 NifLA mutants and plasmids.
A nifLA knock-out mutant was constructed by interrupting the ORF of nifL by insertion of a Sp/Sm-resistance cartridge excised by SmaI from pHP45{Omega} (Prentki & Krisch, 1984 ) into the EcoRV site (Fig. 1) of pTEA24, resulting in pTEA24{Omega}. pTEA24{Omega} was introduced into Azoarcus sp. BH72 by electroporation, yielding double recombinants that had exchanged the interrupted allele with the wild-type allele. In Southern blot analysis, the wild-type showed hybridization of a 14 kb fragment of KpnI-digested genomic DNA with a probe against nifA, while the nifLA mutant, named BHLAO, showed a shift of 2 kb to 16 kb corresponding to the inserted cassette (data not shown). For a NifA- phenotype, strain BHLAO was complemented with pLAFRL, carrying the nifL gene including upstream sequences. pLAFRL was constructed by excising a 2 kb HindIII/BsrBI fragment from pTEA24, and cloning it into HindIII/EcoRV of pBKSII. From this vector the fragment was excised using HindIII/EcoRI and cloned into pLAFR3, yielding pLAFRL. This plasmid was introduced into Azoarcus BHLAO by triparental mating and the resulting strain named BHLAO(pLAFRL). The nifL mutant (BHNLK) was constructed by insertion of a non-polar Km-resistance (aphII) cartridge from pUC4K (AmershamPharmacia Biotech) into the nifL gene. The cartridge was excised by SalI, the overhanging ends blunted using T4 DNA polymerase, and the fragment inserted into the EcoRV site of nifL on pTEA24, yielding pTE24Km. Integration in the correct orientation was confirmed, and the construct was used for electroporation into Azoarcus sp. BH72. A double recombinant (nifL::aphII) confirmed by Southern hybridization (see above) was named BHNLK. For expression studies of nifLA, gusA was transcriptionally fused to the nifL gene by cloning a 1·3 kb PstI fragment from pP13 into the PstI site of pGusKS. This nifL::gusA fusion was excised by HindIII and partial digestion with BamHI and inserted into pLAFR3, yielding pLGus.

ß-Glucuronidase (GUS) assay.
GUS assays were carried out according to Jefferson et al. (1986) . Glucuronidase activity was calculated as (A420x1000)/(t (min)xOD600)=Miller units.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Azoarcus sp. BH72 genomic library clones can complement a nifA mutation in Azotobacter vinelandii and contain genes encoding a NifL and a NifA homologue
In an attempt to find genes homologous to nifA by heterologous complementation of a nifA mutant of Azotobacter vinelandii, a genomic library of Azoarcus sp. strain BH72 was constructed in the cosmid vector pLAFR3 (Tetr); 1440 Tetr clones of E. coli DH5{alpha} were used as cosmid library (insert size approx. 20 kb). After plasmid transfer by triparental mating into Azotobacter vinelandii mutant strain UW1, Nif+ Tetr colonies were selected on N-free BS medium. Out of 10 colonies tested, two types of cosmids with similar restriction fragments were isolated. The cosmids pTEA1 (16 kb insert) and pTEA2 (11 kb insert) allowed N2 fixation of mutant UW1, as did a subcloned 9 kb EcoRI/HindIII fragment (pLAFR24). Sequence analysis of subclones (Fig. 1) revealed two putative ORFs; the deduced amino acid sequence of the first ORF showed 44% and 28% identity to the Azotobacter vinelandii and K. pneumoniae NifL protein, respectively, that of the second ORF showed 61% and 54% identity to Azotobacter vinelandii and K. pneumoniae NifA, respectively. The N-terminal ‘input domains’ of the few known NifL sequences show a significant similarity (amino acid positions 14–130, 67% and 65% with Azotobacter vinelandii and K. pneumoniae, respectively). A PAS domain was detected in this region with statistical significance in SMART (E=1·7x10-4, amino acids 14–84) or PFAM (E=5·4x10-7, amino acids 14–56), as well as an adjacent PAC domain (SMART: amino acids 86–130, E=2·1x10-1; PFAM: amino acids 86–130, E=7·4x10-4) which is proposed to contribute to the PAS domain fold. PAS domains are found in a large family of sensory proteins from all kingdoms including Archaea (Zhulin et al., 1997 ). They are proposed to be the site of FAD-cofactor binding in NifL and may therefore be involved in O2 sensing (Dixon, 1998 ). The C-terminal part of Azoarcus NifL contains five conserved regions in common with other members of the histidine protein kinase family (termed H, N, G1, F and G2 domains). These regions are also present in the Azotobacter vinelandii sequence but are missing (except for G2) in the enterobacterial counterparts. Moreover, Azotobacter vinelandii NifL shows the highly conserved histidine residue of orthodox histidine autokinase transmitter domains (Parkinson & Kofoid, 1992 ) at position 304, which is missing in the Azoarcus and the enterobacterial sequences. However, mutational analysis of Azotobacter vinelandii NifL revealed that phosphorylation of His304 is not essential for any of the known functions of NifL (Woodley & Drummond, 1994 ). Moreover, all attempts to demonstrate NifL autophosphorylation or phosphotransfer to NifA in vitro failed (Austin et al., 1994 ), suggesting that NifL–NifA-mediated signal transduction differs from the classical two-component system paradigm (Dixon, 1998 ).

In Azoarcus NifA, the domain putatively interacting with {sigma}54 and the DNA-binding domains showed a much higher similarity to other NifA proteins than the N-terminal regulatory (or receiver) domains, which are highly variable, possibly indicating various modes of signal transduction in different bacteria. NifA proteins occurring outside the {gamma}-Proteobacteria reveal a characteristic motif of conserved cysteine residues located between the central catalytic domain and the C-terminal DNA-binding domain (Fischer et al., 1988 ) which leads to O2 sensitivity of the protein. This motif is absent in the Azoarcus sequence, which further supports the notion that NifA proteins which operate in concert with NifL proteins are not intrinsically O2 sensitive (Dixon, 1998 ).

Expression of nifLA is affected by O2 and ammonium
The ORFs of nifL and nifA of strain BH72 were closely adjacent (intergenic region 78 bp) (Fig. 1). Only upstream of nifL were motifs characteristic of {sigma}54-dependent promoters (Fig. 2a) corresponding well to the consensus (Merrick, 1992 ). Primer extension analysis with an RNA preparation of an N2-fixing culture of strain BH72 corroborated the transcriptional start site. Using plasmid pS08 as a template for the parallel sequencing reaction, the transcriptional start was localized at minute 145·16, corresponding to the nucleotide marked by an arrow in Fig. 2(a, b) (minute 145·11), which was at position +1 with respect to the -12/-24 motif of the {sigma}54 promoter. Further upstream, motifs similar to the consensus of the NtrC-binding site (GCAC-N7-GTGC) and a motif identical to the consensus of the NifA-binding site (TGT-N10-ACA) were found (Fig. 2a). However, data on nifLA expression in a NifLA mutant (see below) suggest that NifA is not involved in the transcriptional control.



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 2. Analysis of Azoarcus sp. BH72 nifLA transcription. (a) Sequence of the putative promoter region. Putative binding motifs for transcriptional activators (NifA dotted, NtrC dashed), RBS and start codon are depicted. Asterisks mark the bases that are conserved in the promoter. (b) Primer extension analysis localizing the transcriptional start at minute 145·16, corresponding to nucleotide 453 (marked with an arrow). Top, sequencing reaction; bottom, primer extension. (c) Northern blot analysis of 12 µg RNA extracted from cells of Azoarcus sp. BH72 grown aerobically on combined nitrogen (+) or diazotrophically (-). A digoxigenin-labelled probe directed against the nifA gene was used which had been amplified with T3/T7 primers from a 0·5 kb PstI subclone (pP05) of pTEA24 as template. The arrow indicates a transcript of approximately 3·4 kb. (d) Analysis of nifLA cotranscription by RT-PCR. Primers RTnifLAfor and RTnifArev, annealing to nifL or nifA, respectively, were used for RT-PCR with 1 µg RNA isolated from N2-fixing cells. Lane 1, size marker (PstI-digested {lambda} DNA); lane 2, no RNA added; lane 3, reverse transcriptase inactivated (95 °C, 5 min) prior to addition of RNA; lane 4, RNA added without heat inactivation; lane 5, 30 ng chromosomal DNA of strain BH72 as template. (e) Top, analysis of nifA expression by RT-PCR. One microgram of RNA isolated from wild-type BH72 cells grown under ammonium excess (+N) or on N2, and from the nifL::{Omega} mutant BHLAO grown on ammonium, was used for RT-PCR with primers specific for nifA (nifArev3RT, nifAfor1RT, nifArev2RT). Samples were taken after 18 (lane 1), 23 (lane 2) and 28 (lanes 3 and 4) cycles; lane 4, reverse transcriptase heat inactivated as in (d); lane 5, DNA size marker. Products of the expected size (421 bp for nifA and 1242 bp for 16S rRNA) were separated on 1·2% agarose gels and visualized by ethidium bromide staining. Bottom, as a quality control of the RNA, the same RNA preparations were used for 16S rRNA-directed RT-PCR using primers 1401rev, 104f and 1346rev with 10 ng RNA as template; lanes are numbered as above.

 
Northern blot analysis using RNA extracts of strain BH72 revealed only one transcript, approximately 3·4 kb in size, when hybridized with a nifA probe of strain BH72, corresponding to the approximate length of the nifLA coding sequence (Fig. 2c). Cotranscription of nifL and nifA was confirmed by RT-PCR using a primer annealing to nifL and a reverse primer annealing to nifA (Fig. 2d): RT-PCR using DNase-treated RNA of strain BH72 revealed an amplification product of the expected size (422 bp, lane 4) which was obtained also by PCR using chromosomal DNA as a template (lane 5), but not in an RNA-free control (lane 2) or after heat-inactivation of reverse transcriptase (lane 3). In the nifL::{Omega} insertional mutant BHLAO (see below), transcription of nifA was no longer detectable by RT-PCR (Fig. 2e), demonstrating a polar effect on nifA transcription. Cotranscription of nifLA was also found in Enterobacter agglomerans (Siddavattam et al., 1995 ), Azotobacter vinelandii and K. pneumoniae (Dixon, 1998 ). In contrast, in the ß-subgroup member H. seropedicae nifA is not preceded by nifL, is followed by nifB, and is controlled by its own promoter (Souza et al., 1991 ). The latter gene organization is more similar to the organization in {alpha}-Proteobacteria such as Azospirillum (Liang et al., 1991 ).

The nifLA transcript was detectable in aerobically grown cells on combined nitrogen, but was more abundant during N2 fixation (Fig. 2c). This differential expression was confirmed by RT-PCR. RT-PCR amplification products of nifA were more abundant in RNA extracts from N2-fixing cells than in extracts from ammonium-grown cells (Fig. 2e). The use of equal amounts of RNA in both extracts was confirmed by RT-PCR using primers for 16S rRNA (Fig. 2e).

To quantify the expression of the nifLA operon, a transcriptional nifL::gusA fusion was constructed on the broad-host-range vector pLAFR3 (pLGus) which was conjugated into Azoarcus strains. GUS activity of this fusion was tested in the wild-type background as well as the NifLA- background of BHLAO (see below) under various growth conditions 2, 4, 6, 8 and 16 h after incubation (data not shown). Maximum GUS activity was observed at 6 h. During aerobic incubation in both complex (VM) and minimal (SM) medium containing combined nitrogen (including 0·05% NH4Cl) the nifLA operon was expressed (Fig. 3, bars 1 and 2), in accordance with the observed expression of nifLA in the presence of combined nitrogen in Northern blot experiments. O2 limitation increased the expression threefold (Fig. 3, bar 3), while nitrogen limitation (absence of combined nitrogen in N-free SM medium) led to a fivefold increase (Fig. 3, bar 4). Under culture conditions favourable for N2 fixation (1% O2, no combined nitrogen), a six- to sevenfold increase was measured (Fig. 3, bar 5). The expression level (except for N2-fixing conditions) was equal for both genetic backgrounds, suggesting that autoregulation of nifLA does not occur in Azoarcus sp. BH72. Surprisingly, under N2-fixing conditions the nifLA mutant BHLAO reached lower expression levels than the wild-type, which might in part be due to severe growth limitations (the mutant stopped growing after 2–3 h of incubation and is not able to grow on N2 at all; see below).



View larger version (10K):
[in this window]
[in a new window]
 
Fig. 3. Expression of nifLA under various growth conditions: GUS activity of a nifL::gusA fusion in Azoarcus sp. Black columns, wild-type BH72(pLGus); white columns, nifL::{Omega} mutant BHLAO(pLGus). Cells were analysed after 6 h incubation aerobically in complex (VM, 0·1% NH4Cl and yeast extract, 0·3% peptone) (1) or minimal medium (SM+N containing 0·05% NH4Cl, 0·01% yeast extract) (2), microaerobically in minimal medium (3), aerobically in nitrogen-free minimal medium (SM) (4) and microaerobically in N-free minimal medium (5). Error bars show the standard deviation of three independent experiments with 3–5 replicates.

 
While in Azotobacter vinelandii transcription of nifLA is not repressed by ammonium and is not dependent on NtrC or RpoN (Blanco et al., 1993 ), it is NtrC-dependent in K. pneumoniae (Minchin et al., 1988 ) and Enterobacter agglomerans (Siddavattam et al., 1995 ). In Klebsiella and Enterobacter cloacae expression is also strongly enhanced (e.g. approx. tenfold) under anaerobic conditions, mediated by negative supercoiling depending on a DNA gyrase (Dixon et al., 1988 ; Hu et al., 2000 ). In E. cloacae, FNR is additionally involved in this regulation (Hu et al., 2000 ). Azoarcus sp. BH72 appears to be slightly different but shows some similarities to K. pneumoniae in nifLA regulation, since the expression is enhanced both by ammonium deficiency and by microaerobiosis (without an obvious involvement of an FNR box). However, the expression level increased only moderately (three- to fivefold) in strain BH72. Strain BH72 showed no evidence for autoregulation of nifA expression by NifA in a NifLA- mutant, although putative NifA-binding sites are present. This was also reported for H. seropedicae, whose expression through a {sigma}54-dependent promoter is repressed by nitrogen but not by O2 (Souza et al., 2000 ).

The nifA gene product is essential for N2 fixation of Azoarcus sp. strain BH72
To confirm the function of NifLA, a nifL::{Omega} strain (BHLAO) was constructed by marker-exchange mutagenesis, carrying a polar mutation in the nifL gene by insertion of a Sp/Sm-cartridge which also abolished nifA transcription (see above). While growth rates under aerobic conditions in both full (VM) and minimal (SM) medium were not affected in the mutant (not shown), BHLAO was not able to grow on N2 in the absence of combined nitrogen under microaerobic conditions (Fig. 4). This inability to fix N2 was restored when pTEA2 provided the nifLA operon in trans in the mutant BHLAO(pTEA2). Growth of the complemented mutant was comparable to that of the wild-type and a control strain carrying a pLAFR3 vector devoid of insert (Fig. 4). This suggested that a product of the nifLA operon was essential for N2 fixation in Azoarcus sp. strain BH72. A transcriptional fusion of Azoarcus sp. nifH with gusA (pEGN3.1; Egener et al., 1999 ) was integrated into the chromosome of mutant BHLAO by single recombination, resulting in strain BHLAONG. In comparison to the wild-type BHNG3.1 (Egener et al., 1999 ), transcriptional activation of the nifH::gusA fusion under conditions for N2 fixation was not observed in BHLAONG (745±69 or 20±9 Miller units, respectively), implying that the nifLA operon encodes the transcriptional activator of nifHDK genes in Azoarcus sp. BH72. To obtain a NifA- phenotype, the double mutant BHLAO was complemented with a 2 kb HindIII/BsrI fragment of pTEA24 carrying only nifL including upstream regions (plasmid pLAFRL). This strain BHLAO(pLAFRL) was not capable of N2 fixation (data not shown), similar to mutant BHLAO, indicating that NifA and not NifL was essential for diazotrophy as a transcriptional activator. These results also confirmed that nifLA are cotranscribed in strain BH72.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 4. Role of Azoarcus sp. nifLA for growth on N2. Growth on N2 of wild-type Azoarcus sp. BH72 ({blacksquare}), of wild-type carrying the vector control [BH72(pLAFR3), {blacktriangledown}], of the niL::{Omega} mutant BHLAO ({square}) and of BHLAO complemented with nifLA [BHLAO(pTEA2), {circ}]. Error bars show the standard deviations from two replicates in three independent experiments.

 
NifL is required for control of nifH expression in response to O2 and combined nitrogen
To construct a nifL mutant, a non-polar Km-resistance cartridge was inserted into the EcoRV site disrupting the nifL gene, resulting in mutant strain BHNLK after allelic exchange. This mutant strain had similar levels of nifA mRNA expression as the wild-type when grown microaerobically on SM medium supplemented with 10 mM NH4Cl, as shown by RT-PCR (Fig. 5a).Thus, insertion of the Km-resistance cartridge did not lead to unusually high nifA mRNA levels under repressing conditions.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 5. Nitrogenase gene expression and acetylene reduction of a nifL mutant. (a) Top, comparison of nifA mRNA abundance by RT-PCR (nifA) in wild-type BH72 and the non-polar nifL::aphII mutant BHNLK grown on ammonium under microaerobic conditions. Reverse transcription was carried out on 1 µg RNA, followed by PCR with 18–26 cycles; samples were taken after 18 (lane 1), 20 (lane 2), 22 (lane 3), 24 (lane 4) and 26 (lane 5) PCR cycles; lane 6, control reaction (26 cycles) without reverse transcription (heat inactivation). Products of the expected size (421 bp for nifA and 1242 bp for 16S rRNA) were separated on 1·2% agarose gels and visualized by ethidium bromide staining. Bottom, as a control, 16S rRNA-specific RT-PCR was carried out on 10 ng RNA. Lanes as for nifA, except for lane 6 (28 cycles) and lane 7 (control reaction at 28 cycles without reverse transcription). Middle lane, PstI-digested lambda DNA as size marker. All RT-PCR reactions were independently performed three times and gave consistently similar results. (b) Expression of a nifH::gusA fusion in the non-polar nifL mutant BHNLKnifH::gusA in response to O2 concentrations on N-free medium; (c, d) acetylene reduction (c) and GUS activity (d) in response to NH4Cl concentration under microaerobic conditions (headspace 1% O2). Error bars indicate the standard deviation from three replicates in two independent experiments.

 
To study the influence of NifL on nif gene expression, a nifH::gusA reporter gene fusion (Egener et al., 1999 ) was additionally introduced into the nifL mutant by single homologous recombination of pEGN3.1, resulting in strain BHNLKnifH::gusA. Wild-type Azoarcus sp. strain BH72 showed an optimal nifH::gusA expression under microaerobic conditions of 1% O2 in the headspace, expression being completely abolished at 4% O2 (reduction of GUS expression by a factor of 40) (Egener et al., 1999 ). In contrast, the nifL mutant did not show this response but expressed nifH to similar levels from 1% to 8% O2 in the headspace (Fig. 5b), indicating that expression was not modulated in response to different O2 levels in the absence of NifL. The influence of ammonium on nifH gene expression and nitrogenase activity was measured in BHNLKnifH::gusA incubated microaerobically with different NH4Cl concentrations. The activity of nitrogenase decreased at ammonium concentrations above 0·5 mM in the nifL mutant (Fig. 5c) as in the wild-type. This was probably due to a rapid inactivation (‘switch-off’) of nitrogenase by ammonium, which has recently been demonstrated for strain BH72 (Egener et al., 2001 ). In contrast, the nifH gene expression remained at a high level and even showed a significant increase with rising ammonium concentrations (Fig. 5d), while in the wild-type strain nifH::gusA expression is abolished at 0·5 mM NH4Cl (reduction by a factor of approx. 40-fold) (Egener et al., 1999 ). This increase might be due to a more rapid growth of the strain in the presence of ammonium. Thus, the NifA protein alone was not inactivated by ammonium in the absence of NifL. It is unlikely that differences in nifA transcription accounted for the observed derepression of nifH under permissive conditions: expression levels of nifA under non-permissive conditions are relatively high (albeit elevated under nitrogen-limitation) in the wild-type, and no change of nifA expression was detected in the nonpolar nifL mutant. Nitrogen-responsiveness of the NifLA regulatory system in Azoarcus sp. might be mediated by PII-like proteins as postulated for members of the {gamma}-Proteobacteria (He et al., 1998 ; Jack et al., 1999 ; Little et al., 2000 ) (unpublished observations). Unlike in most proteobacteria, three instead of two paralogues of PII-like proteins were found in Azoarcus sp. BH72 (Martin et al., 2000 ).

Concluding remarks
Our results have demonstrated that with respect to the function of NifA in signal transduction, Azoarcus sp. strain BH72 is more similar to diazotrophs of the {gamma}-subgroup than to the closer relative H. seropedicae, both genera being members of the ß-subgroup. As in bacteria of the {alpha}-subgroup, in H. seropedicae NifA activity is self-modulated: NifA ativity is sensitive to combined nitrogen and O2 (Souza et al., 1999 ). In contrast, in Azoarcus sp. BH72 NifA by itself was not able to modulate nifH transcription in response to ammonium or O2 when nifL was inactivated. This suggests that O2 signalling or sensing occurs through the NifL protein as described for diazotrophs of the {gamma}-subgroup (Dixon, 1998 ), reflecting the phylogenetic relationship of the ß- and {gamma}-subgroups according to 16S rDNA analysis. Interestingly, in H. seropedicae also the iron protein of nitrogenase is located in a clade of {alpha}-proteobacterial proteins according to phylogenetic analysis, leading to the speculation that the structural nif genes in H. seropedicae might have been gained by lateral gene transfer (Hurek et al., 1997a ). The similarities of details of transcriptional regulation of nif genes to {alpha}-subgroup diazotrophs support the view that the entire N2-fixing apparatus in this bacterium might have been gained by lateral gene transfer.


   ACKNOWLEDGEMENTS
 
We thank Christina Kennedy, University of Arizona, Tuscon, for kindly providing Azotobacter vinelandii mutants, and Thomas Hurek, Cooperation Laboratory Max-Planck-Institute for Marine Microbiology/University of Bremen, for valuable discussions. This work was supported by a grant of the Deutsche Forschungsgemeinschaft to B.R.-H. (Re 756/5-2 and 5-3).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403-410.[Medline]

Arcondéguy, T., Jack, R. & Merrick, M. (2001). PII signal transduction proteins, pivotal players in microbial nitrogen control. Microbiol Mol Biol Rev 65, 80-105.[Abstract/Free Full Text]

Arsène, F., Kaminski, P. A. & Elmerich, C. (1996). Modulation of NifA activity by PII in Azospirillum brasilense: evidence for a regulatory role of the NifA N-terminal domain. J Bacteriol 178, 4830-4838.[Abstract]

Arsène, F., Kaminski, P. A. & Elmerich, C. (1999). Control of Azospirillum brasilense NifA activity by PII: effect of replacing Tyr residues of the NifA N-terminal domain on NifA activity. FEMS Microbiol Lett 179, 339-343.[Medline]

Austin, S., Buck, M., Cannon, W., Eydmann, T. & Dixon, R. (1994). Purification and in vitro activities of the native nitrogen fixation control proteins NifA and NifL. J Bacteriol 176, 3460-3465.[Abstract]

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (editors) (1987). Current Protocols in Molecular Biology. New York: Wiley.

Blanco, G., Drummond, M., Woodley, P. & Kennedy, C. (1993). Sequence and molecular analysis of the nifL gene of Azotobacter vinelandii. Mol Microbiol 9, 869-879.[Medline]

Dixon, R. (1998). The oxygen-responsive NifL–NifA complex: a novel two-component regulatory system controlling nitrogenase synthesis in {gamma}-Proteobacteria. Arch Microbiol 169, 371-380.[Medline]

Dixon, R. A., Henderson, N. C. & Austin, S. (1988). DNA supercoiling and aerobic regulation of transcription from the Klebsiella pneumoniae nifLA promoter. Nucleic Acids Res 16, 9933-9946.[Abstract]

Drummond, M., Clements, J., Merrick, M. & Dixon, R. (1983). Positive control and autogenous regulation of the nifLA promoter in Klebsiella pneumoniae. Nature 301, 302-307.[Medline]

Egener, T., Hurek, T. & Reinhold-Hurek, B. (1999). Endophytic expression of nif genes of Azoarcus sp. strain BH72 in rice roots. Mol Plant–Microbe Interact 12, 813-819.

Egener, T., Martin, D. E., Sarkar, A. & Reinhold-Hurek, B. (2001). Role of a ferrodoxin gene cotranscribed with the nifHDK operon in N2 fixation and nitrogenase ‘switch off’ of Azoarcus sp. strain BH72. J Bacteriol 183, 3752-3760.[Abstract/Free Full Text]

Fischer, H. M., Bruderer, T. & Hennecke, H. (1988). Essential and non-essential domains in the Bradyrhizobium japonicum NifA protein: identification of indispensable cysteine residues potentially involved in redox reactivity and/or metal binding. Nucleic Acids Res 16, 2207-2224.[Abstract]

Govantes, F., Molina-Lopez, J. A. & Santero, E. (1996). Mechanism of coordinated synthesis of the antagonistic regulatory proteins NifL and NifA of Klebsiella pneumoniae. J Bacteriol 178, 6817-6823.[Abstract]

Hanahan, D. (1983). Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166, 557-580.[Medline]

He, L., Soupene, E., Ninfa, A. & Kustu, S. (1998). Physiological role for the GlnK protein of enteric bacteria: relief of NifL inhibition under nitrogen-limiting conditions. J Bacteriol 180, 6661-6667.[Abstract/Free Full Text]

Hill, S., Austin, S., Eydmann, T., Jones, T. & Dixon, R. (1996). Azotobacter vinelandii NifL is a flavoprotein that modulates transcriptional activation of nitrogen-fixation genes via a redox-sensitive switch. Proc Natl Acad Sci USA 93, 2143-2148.[Abstract/Free Full Text]

Hu, B., Zhu, J., Shen, S. C. & Yu, G. Q. (2000). A promoter region binding protein and DNA gyrase regulate anaerobic transcription of nifLA in Enterobacter cloacae. J Bacteriol 182, 3920-3923.[Abstract/Free Full Text]

Hurek, T., Reinhold, B., Fendrik, I. & Niemann, E. G. (1987). Root-zone-specific oxygen tolerance of Azospirillum spp. and diazotrophic rods closely associated with Kallar grass. Appl Environ Microbiol 53, 163-169.

Hurek, T., Burggraf, S., Woese, C. R. & Reinhold-Hurek, B. (1993). 16S rRNA-targeted polymerase chain reaction and oligonucleotide hybridization to screen for Azoarcus spp., grass-associated diazotrophs. Appl Environ Microbiol 59, 3816-3824.[Abstract]

Hurek, T., Reinhold-Hurek, B., Turner, G. L. & Bergersen, F. J. (1994a). Augmented rates of respiration and efficient nitrogen fixation at nanomolar concentrations of dissolved O2 in hyperinduced Azoarcus sp. strain BH72. J Bacteriol 176, 4726-4733.[Abstract]

Hurek, T., Reinhold-Hurek, B., Van Montagu, M. & Kellenberger, E. (1994b). Root colonization and systemic spreading of Azoarcus sp. strain BH72 in grasses. J Bacteriol 176, 1913-1923.[Abstract]

Hurek, T., Van Montagu, M., Kellenberger, E. & Reinhold-Hurek, B. (1995). Induction of complex intracytoplasmic membranes related to nitrogen fixation in Azoarcus sp. BH72. Mol Microbiol 18, 225-236.[Medline]

Hurek, T., Wagner, B. & Reinhold-Hurek, B. (1997a). Identification of N2-fixing plant- and fungus-associated Azoarcus species by PCR-based genomic fingerprints. Appl Environ Microbiol 63, 4331-4339.[Abstract]

Hurek, T., Egener, T. & Reinhold-Hurek, B. (1997b). Divergence in nitrogenases of Azoarcus spp., Proteobacteria of the ß-subclass. J Bacteriol 179, 4172-4178.[Abstract]

Jack, R., De Zamaroczy, M. & Merrick, M. (1999). The signal transduction protein GlnK is required for NifL-dependent nitrogen control of nif gene expression in Klebsiella pneumoniae. J Bacteriol 181, 1156-1162.[Abstract/Free Full Text]

Jefferson, R. A., Burgess, S. M. & Hirsh, D. (1986). ß-Glucuronidase from Escherichia coli as a gene-fusion marker. Proc Natl Acad Sci USA 83, 8447-8451.[Abstract]

Karg, T. & Reinhold-Hurek, B. (1996). Global changes in protein composition of N2-fixing Azoarcus sp. strain BH72 upon diazosome formation. J Bacteriol 178, 5748-5754.[Abstract/Free Full Text]

Liang, Y. Y., Kaminski, P. A. & Elmerich, C. (1991). Identification of a nifA-like regulatory gene of Azospirillum brasilense Sp7 expressed under conditions of nitrogen fixation and in the presence of air and ammonia. Mol Microbiol 5, 2735-2744.[Medline]

Little, R., Reyes-Ramirez, F., Zhang, Y., van Heeswijk, W. C. & Dixon, R. (2000). Signal transduction to the Azotobacter vinelandii NifL–NifA regulatory system is influenced directly by interaction with 2-oxoglutarate and the PII regulatory protein. EMBO J 19, 6041-6050.[Abstract/Free Full Text]

Martin, D., Hurek, T. & Reinhold-Hurek, B. (2000). Occurrence of three PII-like signal transmitter proteins in the diazotroph Azoarcus sp. BH72. Mol Microbiol 38, 276-288.[Medline]

Merrick, M. (1992). Regulation of nitrogen fixation genes in free-living and symbiotic bacteria. In Biological Nitrogen Fixation , pp. 835-876. Edited by G. Stacey, R. Burris & H. Evans. New York:Chapman & Hall.

Minchin, S. D., Austin, S. & Dixon, R. A. (1988). The role of activator binding sites in transcriptional control of the divergently transcribed nifF and nifLA promoters from Klebsiella pneumoniae. Mol Microbiol 2, 433-442.[Medline]

Newton, J. W., Wilson, P. W. & Burris, R. H. (1953). Direct demonstration of ammonia as an intermediate in nitrogen fixation by Azotobacter. J Biol Chem 204, 445-451.[Free Full Text]

Ninfa, A. J. & Atkinson, M. R. (2000). PII signal transduction proteins. Trends Microbiol 8, 172-189.[Medline]

Page, W. J. & Sadoff, H. L. (1976). Physiological factors affecting transformation of Azotobacter vinelandii. J Bacteriol 125, 1080-1087.[Medline]

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

Prentki, P. & Krisch, H. M. (1984). In vitro insertional mutagenesis with a selectable DNA fragment. Gene 29, 303-314.[Medline]

Reinhold, B., Hurek, T. & Fendrik, I. (1985). Strain-specific chemotaxis of Azospirillum spp. J Bacteriol 162, 190-195.[Medline]

Reinhold, B., Hurek, T., Niemann, E.-G. & Fendrik, I. (1986). Close association of Azospirillum and diazotrophic rods with different root zones of Kallar grass. Appl Environ Microbiol 52, 520-526.

Reinhold-Hurek, B., Hurek, T., Claeyssens, M. & Van, M. M. (1993a). Cloning, expression in Escherichia coli, and characterization of cellulolytic enzymes of Azoarcus sp., a root-invading diazotroph. J Bacteriol 175, 7056-7065.[Abstract]

Reinhold-Hurek, B., Hurek, T., Gillis, M., Hoste, B., Vancanneyt, M., Kersters, K. & De Ley, J. (1993b). Azoarcus gen. nov., nitrogen-fixing proteobacteria associated with roots of Kallar grass (Leptochloa fusca (L.) Kunth.) and description of two species Azoarcus indigens sp. nov. and Azoarcus communis sp. nov. Int J Syst Bacteriol 43, 574-584.

Schultz, J., Milpetz, F., Bork, P. & Ponting, C. P. (1998). SMART, a simple modular architecture research tool: identification of signaling domains. Proc Natl Acad Sci USA 95, 5857-5864.[Abstract/Free Full Text]

Shah, V. K., Davis, I. C., Gordon, J. K., Orme-Johnson, W. H. & Brill, W. J. (1973). Nitrogenaseless mutants of Azotobacter vinelandii: activities, cross-reactions and EPR spectra. Biochim Biophys Acta 292, 246-255.[Medline]

Siddavattam, D., Steibl, H.-D., Kreutzer, R. & Klingmüller, W. (1995). Regulation of nif gene expression in Enterobacter agglomerans: nucleotide sequence of the nifLA operon and influence of temperature and ammonium on its transcription. Mol Gen Genet 249, 629-636.[Medline]

Sonnhammer, E. L., Eddy, S. R. & Durbin, R. (1997). Pfam: a comprehensive database of protein domain families based on seed alignments. Proteins 28, 405-420.[Medline]

Souza, E. M., Funayama, S., Rigo, L. U., Yates, M. G. & Pedrosa, F. O. (1991). Sequence and structural organization of a nifA-like gene and part of a nifB-like gene of Herbaspirillum seropedicae strain Z78. J Gen Microbiol 137, 1511-1522.[Medline]

Souza, E. M., Pedrosa, F. O., Drummond, M., Rigo, L. U. & Yates, M. G. (1999). Control of Herbaspirillum seropedicae NifA activity by ammonium ions and oxygen. J Bacteriol 181, 681-684.[Abstract/Free Full Text]

Souza, E. M., Pedrosa, F. O., Rigo, L. U., Machado, H. B. & Yates, M. G. (2000). Expression of the nifA gene of Herbaspirillum seropedicae: role of the NtrC and NifA binding sites and of the -24/-12 promoter element. Microbiology 146, 1407-1418.[Abstract/Free Full Text]

Staskawicz, B., Dahlbeck, D., Keen, N. & Napoli, C. (1987). Molecular characterisation of cloned avirulence genes from race 0 and race 1 of Pseudomonas syringae pv. glycinea. J Bacteriol 169, 5789-5794.[Medline]

Woodley, P. & Drummond, M. (1994). Redundancy of the conserved His residue in Azotobacter vinelandii NifL, a histidine autokinase homologue which regulates transcription of nitrogen fixation genes. Mol Microbiol 13, 619-626.[Medline]

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.[Medline]

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.[Medline]

Received 3 January 2002; revised 14 June 2002; accepted 20 June 2002.