Azorhizobium caulinodans electron-transferring flavoprotein N electrochemically couples pyruvate dehydrogenase complex activity to N2 fixation

John D. Scott and Robert A. Ludwig

Department of Molecular, Cellular and Developmental Biology, Sinsheimer Laboratories, University of California, Santa Cruz, CA 95064, USA

Correspondence
Robert A. Ludwig
ludwig{at}biology.ucsc.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Azorhizobium caulinodans thermolabile point mutants unable to fix N2 at 42 °C were isolated and mapped to three, unlinked loci; from complementation tests, several mutants were assigned to the fixABCX locus. Of these, two independent fixB mutants carried missense substitutions in the product electron-transferring flavoprotein N (ETFN) {alpha}-subunit. Both thermolabile missense variants Y238H and D229G mapped to the ETFN{alpha} interdomain linker. Unlinked thermostable suppressors of these two fixB missense mutants were identified and mapped to the lpdA gene, encoding dihydrolipoamide dehydrogenase (LpDH), immediately distal to the pdhABC genes, which collectively encode the pyruvate dehydrogenase (PDH) complex. These two suppressor alleles encoded LpDH NAD-binding domain missense mutants G187S and E210G. Crude cell extracts of these fixB lpdA double mutants showed 60–70 % of the wild-type PDH activity; neither fixB lpdA double mutant strain exhibited any growth phenotype at the restrictive or the permissive temperature. The genetic interaction between two combinations of lpdA and fixB missense alleles implies a physical interaction of their respective products, LpDH and ETFN. Presumably, this interaction electrochemically couples LpDH as the electron donor to ETFN as the electron acceptor, allowing PDH complex activity (pyruvate oxidation) to drive soluble electron transport via ETFN to N2, which acts as the terminal electron acceptor. If so, then, the A. caulinodans PDH complex activity sustains N2 fixation both as the driving force for oxidative phosphorylation and as the metabolic electron donor.


Abbreviations: ETF, electron-transferring flavoprotein; LpDH, dihydrolipoamide dehydrogenase; PDH, pyruvate dehydrogenase; Nifts, N2 fixation temperature-sensitive

The nucleotide sequence for the Azorhizobium caulinodans dihydrolipoamide dehydrogenase gene (lpdA) described in this article has been deposited in GenBank under accession number AY331184.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Among diazotrophic and endosymbiotic bacteria, N2 fixation activity is notably thermolabile, typically showing little or no activity at or above 37 °C. Moreover, among legumes indigenous to temperate climates, symbiotic N2 fixation activity is often impaired at or above 30 °C. Yet legume endosymbionts of the family Rhizobiaceae include Azorhizobium caulinodans, whose N2 fixation activities both in culture and in symbiosis are remarkably thermostable in this temperature range and above. To what is owed this highly unusual thermostability?

To study this question, we isolated and characterized A. caulinodans thermolabile mutants lacking N2 fixation activity at 42 °C. Both weak and null thermolabile alleles were identified, mapped and found to comprise distinct categories. As the A. caulinodans loci encoding N2 fixation genes had been previously identified, studied and mapped (Donald et al., 1986), we were able to draw functional inferences from these thermolabile mutants.

In A. caulinodans, N2 fixation locus-2 includes a tight cluster of four genes: the fixAB genes, encoding an electron-transferring flavoprotein (ETFN), the fixC gene, encoding a presumed ETFN oxidoreductase, and a fixX gene, encoding a [4Fe–4S]ferredoxin. While the ETF family is widely dispersed among bacteria and eukaryotes, all ETFs are highly conserved {alpha}/{beta}-heterodimeric proteins (Finocchiaro et al., 1988; O'Neill et al., 1998). Structurally, the human mitochondrial ETF archetype comprises three folded domains; one stably binds flavin adenine dinucleotide (FAD), a second stably binds 5'-adenylic acid (5'-AMP) and a third, a flexible hinge, facilitates generic flavoprotein–flavoprotein interactions. Functionally, in bacteria and eukaryotes, ETF, as a soluble one- or two-electron carrier, couples various flavoprotein dehydrogenases, one acting oxidatively, the other reductively, via transient formation of alternately ETF-oxidized and ETF-reduced binary complexes. While inherent rates of ETF-mediated soluble electron transfer processes are slow, ETF activity nevertheless facilitates the overall cellular oxidation–reduction balance. In eukaryotic mitochondria, conceptually, ETF couples to membrane respiration various oxidations carried out by matrix (soluble) flavoprotein dehydrogenases. Itself a matrix protein, ETF primarily serves as an electron acceptor for fatty acid oxidation, re-oxidizing a set of four (long-, medium-, short- and branched-chain) fatty acyl-CoA dehydrogenases. In turn, reduced mitochondrial ETF donates electrons to ETF : ubiquinone oxidoreductase (ETF-QO), a peripheral mitochondrial membrane activity, and thence to membrane respiration (Goodman et al., 1994).

In the Rhizobiaceae, a family of microaerophilic bacteria, a specialized ETFN facilitates both symbiotic (Corbin et al., 1983; Earl et al., 1987) and free-living (Donald et al., 1986) N2 fixation by acting as the electron-transfer intermediary. In electrochemical terms, N2 fixation is a highly reductive process. The dinitrogenase complex, the reactive core of the process, consumes eight electron-equivalents per mole of N2 substrate fixed as (two equivalents of) ammonium and (one equivalent of) H2. Dinitrogenase complex activity is also enormously ATP-consumptive in vivo, requiring 16–24 ATP equivalents per N2 equivalent fixed. Given this complexity, the precise role for ETFN-mediated electron transfer in support of N2 fixation has remained uncertain: from what does it receive and to what does it donate electrons?

We report here a gene-for-gene interaction between A. caulinodans ETFN and the pyruvate dehydrogenase (PDH) complex, specifically its dihydrolipoamide dehydrogenase (LpDH) component. The PDH complex, which oxidizes pyruvate to acetyl-CoA, includes three catalytic centres: PDH (E1), dihydrolipoamide S-acetyltransferase (E2) and LpDH (E3). LpDH is a member of the pyridine-nucleotide disulfide oxidoreductase superfamily and carries a single, tightly bound FAD prosthetic group. In bacteria, typically, LpDH is encoded by several paralogous genes, each of which map to tightly linked gene sets encoding the various 2-oxoacid (pyruvate, 2-oxoglutarate and 2-oxoisovalerate) dehydrogenase complexes.

In bacteria and eukaryotes, NAD+ normally serves the PDH complex as the electron acceptor under aerobic physiological conditions. However, as we report here, in microaerobic A. caulinodans cultures fixing N2, ETFN also re-oxidizes the PDH complex. By inference, ETFN probably shuttles reducing equivalents via soluble electron carriers to the dinitrogenase complex, sustaining N2 fixation. Thus, PDH complex activity facilitates N2 fixation in two distinct ways: (i) with NAD+ as an oxidant, it drives membrane-linked respiration for ATP synthesis using O2 as the terminal electron acceptor, and (ii) with ETFN as an oxidant, it drives soluble electron transport using N2 as the terminal electron acceptor.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, culture methods and mutagenesis.
An A. caulinodans library carrying random, single, vector-insertion (Vi) mutants (Donald et al., 1985; Donald et al., 1986) was screened for thermolabile N2 fixation-defective strains at 42 °C as the restrictive and 30 °C as the permissive condition in defined N2 fixation (NIF) salts medium (0·4 % potassium succinate, 50 mM potassium phosphate pH 6·4, 1·25 mM magnesium sulfate, 300 µM potassium nicotinate, 0·5 mM calcium chloride, 1 µg NaMoO4 ml-1, 1 µg FeCl3 ml-1). For outgrowth under permissive conditions, A. caulinodans strains were propagated in defined minimal medium (ORS-MM: 0·4 % potassium succinate, 10 mM ammonium sulfate, 10 mM potassium phosphate pH 6·3, 1 mM magnesium sulfate, 0·5 mM calcium chloride, 16 µg nicotinate ml-1, 1·0 µg pantothenate ml-1, 0·2 µg biotin ml-1). For liquid batch culture experiments, test strains were first cultured in rich GYPC medium (0·4 % D-glucose, 0·2 % yeast extract, 10 mM potassium phosphate pH 6·3, 0·2 % salt-free casein acid hydrolysate). Culture growth was monitored spectrophotometrically (OD600). To measure N2-dependent growth, A. caulinodans strains were first cultured in liquid ORS-MM under air sparge to late-exponential phase (OD600 0·7). Bacterial cells were recovered by centrifugation, washed twice and diluted to 1x108 cells ml-1 (OD600 0·05) in NIF medium; cell samples were placed in stoppered serum vials and again cultured at 30 °C under sparge with O2/CO2/N2 (1 % : 1 % : 98 %). For N2-dependent colony growth tests, NIF medium was solidified with acid-washed (1 M HCl) agarose and plates were incubated in sealed jars at 30 °C under continuous sparging with O2/CO2 (0·1 % : 1·0 %), balance Ar; bacterial colonies were examined after 7 days incubation and again after 14 days incubation.

Thermolabile N2 fixation-defective mutants were also isolated after ethyl methanesulfonate (EMS) mutagenesis, followed by repetitive enrichment. A. caulinodans 57100 (wild-type) was cultured in GYPC medium at 30 °C to mid-exponential phase (5x108 cells ml-1; OD600 0·25). Cells were recovered by centrifugation, washed with EB (0·075 M potassium phosphate pH 6·3, 1·5 mM ammonium sulfate, 0·8 mM magnesium sulfate, 0·05 mM calcium chloride), recovered by centrifugation and resuspended four-fold concentrated in EB. EMS was added to 0·25 % (w/v), and cultures were incubated with vigorous shaking for 30 min at 30 °C (~1 % survival). Mutagenized cells were recovered by centrifugation, washed with GYPC, centrifuged again, resuspended 100-fold-diluted in GYPC and cultured overnight at 30 °C with vigorous shaking. To screen for N2 fixation-defective missense mutants, washed, EMS-mutagenized, outgrown cell samples (1x108 cells ml-1; OD600 0·05) were induced for 4 h at 42 °C as described above for isolation of null mutants. Following this, Timentin® (0·1 mg ml-1), a proprietary mixture of ticarcillin and {beta}-clavulanate, was added and cultures were incubated for 6 h at 42 °C; continuous sparging with a mixture of O2/CO2/N2 (1 % : 1 % : 98 %) was maintained. In control experiments, viable cell counts showed ~0·1 % survival, measured as colony-forming ability on GYPC, after this treatment. Surviving cells were recovered by centrifugation and subcultured overnight at 30 °C in ORS-MM. Enrichments were repeated for a total of four cycles. Surviving cells were first plated on ORS-MM at 42 °C and subsequently replica-plated on NIF medium plates incubated at either 30 or 42 °C for 14 days in sealed jars under continuous sparging with a mixture of O2/CO2/N2 (1 % : 1 % : 98 %); temperature-sensitive mutants were identified and characterized further (see Results).

Molecular cloning and DNA sequencing.
The nucleotide sequence for the A. caulinodans lpdA gene located on recombinant plasmid pSKC42, derived from recombinant {lambda}pdhC4, was determined by standard techniques (Sambrook et al., 1989). Identification of the gene product was made by BLASTP search versus the san diego supercomputer center non-redundant protein database and confirmed by multiple PIMA algorithm alignments (Smith & Smith, 1990).

Dinitrogenase assay by acetylene reduction activity.
A. caulinodans N2 fixation was measured in culture by acetylene reduction and N2-dependent growth. Cells were grown to late-exponential phase in ORS-MM supplemented with 0·1 mM nicotinate, harvested, washed twice, resuspended (4x108 cells ml-1) in NIF medium, placed in stoppered serum vials and sparged with O2/CO2/Ar (3 % : 1 % : 96 %) for 8 h at 30 °C. Vials were then injected with 0·2 atm acetylene (substrate) freshly generated by hydration of calcium carbide, and assayed for ethylene production versus time using GC (Donald et al., 1985).

PDH activity.
Mid-exponential phase cultures in ORS-MM were isolated and cell-free extracts were prepared by ultrasonication. Extracts were then cleared by centrifugation (1 h at 100 000 g) and assayed immediately or, for determination of kinetic constants, further purified by immunoprecipitation. Cleared extracts were made by addition of 1 % Nonidet P40 (NP40), 0·5 % sodium deoxycholate and 10 mM PMSF, and treated for 1 h at 4 °C with 1 % polyclonal antisera prepared against purified Mycoplasma pneumoniae PDH-B (Dallo et al., 2002); 5 % protein-A agarose suspension (Boehringer Mannheim) was then added and samples were incubated for a further 3 h at 4 °C with slow tumbling. Samples were precipitated by centrifugation (1 min at 15 000 g), washed/precipitated twice for 20 min at 4 °C in 1 vol. (original extract) of a mixture of 50 mM Tris/Cl pH 7·4, 150 mM NaCl, 1 % NP40, 0·5 % sodium deoxycholate and 1mM PMSF, and again precipitated by centrifugation. Purified PDH was eluted by resuspension and overnight incubation with 50 mM Tris/Cl pH 7·4, 0·5 M NaCl, 0·1 % NP40 and 0·05 % sodium deoxycholate at 4 °C with slow tumbling. After centrifugation (1 min at 15 000 g) at 4 °C, the supernatant containing eluted PDH was dialysed against a mixture of 0·1 M Tris/Cl pH 7·5, 10 mM MgCl2, 0·1 mM CoA and 1 mM DTT at 4 °C. Protein concentrations were measured by standard Bradford assays. PDH activity at 30 °C was inferred as pyruvate-dependent reduction of NAD+ measured as spectrophotometric absorbance at 339 nm. Reactions comprised 0·1 M Tris/Cl pH 7·5, 10 mM MgCl2, 0·12 mM CoA, 6 mM DTT, 1·2 mM NAD+ and supernatant fractions of cell-free extracts (500 µg total protein), and were initiated with 10 mM pyruvate. Pyruvate-dependent cytochrome reduction was assayed by ferricyanide reduction. Batch cultures were grown on ORS-MM supplemented with 0·2 % potassium succinate, 0·1 % monosodium glutamate and 0·1 mM potassium nicotinate; early-exponential phase cultures were harvested and washed; cell pellets were resuspended and lysed by ultrasonication. Crude cell-free extracts were then used to measure pyruvate-dependent reduction of ferricyanide at 30 °C measured as spectrophotometric absorbance at 420 nm. Reactions contained 0·1 M Tris/Cl pH 7·5, 10 mM MgCl2, 0·07 mM potassium ferricyanide and added crude cell-free extract (100 µg total protein), and were initiated with 20 mM pyruvate.

Sesbania rostrata nodulation tests.
S. rostrata seedlings were germinated aseptically and grown on sterile, defined medium under nitrogen limitation (Kwon & Beevers, 1992). Three-week-old plants, some 50 cm in height, were inoculated with the desired A. caulinodans strain between first and second stem internodes, for which region stem-nodule development is synchronized (Donald et al., 1986). Starting 6 days post-inoculation, maturing stem-nodules were excised and tested for N2 fixation activity by acetylene reduction. Thereafter, nodules were recovered, crushed and homogenized (Polytron); cell-free supernatants were analysed for leghaemoglobin measured as spectrophotometric absorbance at 540 nm. All developmental nodulation tests were done in triplicate.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A. caulinodans fixes N2 at high temperature (42 °C) yet lacks novel genes specifically conferring thermostable N2 fixation
Mature S. rostrata plants nodulated with A. caulinodans 57100 (wild-type isolate) and cultivated in defined nutrient salts limited for available-nitrogen reproducibly grew faster at 37 °C than at 30 °C when given saturating light intensities and a 12 h photoperiod. Enhanced growth required inoculation of seedlings with A. caulinodans 57100. In pure cultures, A. caulinodans 57100 dinitrogenase complex rates were undiminished at 37 °C when compared to 30 °C; at 42 °C, rates were approximately 40 % as measured by acetylene reduction. When A. caulinodans 57100 was cultured with defined media in which N2 was the predominant nitrogen source, growth yields corroborated with dinitrogenase activities (results not presented). To help understand (this unusual) thermostable N2 fixation, two searches were carried out in parallel, and screening was done in several steps. In the first search, a strain library of A. caulinodans 57100 derivatives carrying single-copy, random Vi2021 vector-insertions was used (Donald et al., 1985); in the second search, a strain library of chemically mutagenized A. caulinodans 57100 derivatives was used.

In both searches, batch cultures were replicated onto solid, defined, nitrogen-limited (NIF) medium (see Methods) and duplicate plates were incubated for 5 days at 30 or 42 °C. Duplicate plates were then superimposed and opaque (dense) colonies specific to plates incubated at 30 °C were identified. Candidate strains were isolated, purified and re-tested for growth both in liquid and on solid defined media. Six, nominally independent candidates from the first (insertion mutant) search, strains 60501–60506, were individually retested for growth on solid NIF medium; while all strains showed absolutely no growth at the restrictive temperature, they showed only limited growth at the permissive temperature. At the restrictive temperature, all six candidates grew indistinguishably from the wild-type on NIF medium supplemented with 10 mM ammonium. Despite repeated searches of the random insertion mutant library, no candidates that both retained wild-type N2-dependent growth at the permissive temperature and completely lacked growth at the restrictive temperature were identified.

Strains 60501–60506 were subjected to direct molecular (vector-insertion) cloning as described previously (Donald et al., 1985), and recombinant plasmids for each mutant were obtained (Table 1; see Methods). These plasmids (pDRN60501 to pDRN60506) were used as DNA hybridization probes against a battery of recombinant {lambda} phages carrying wild-type nif loci (Donald et al., 1986). All six plasmids hybridized generally with phage {lambda}nif3 DNA and specifically to a 3·0 kbp BglII DNA fragment carrying nifK and nifE DNA homology. DNA sequencing analysis using as primer a specific IS50 deoxyoligonucleotide sequence of the insertion element allowed delineation of precise insertion points in the mutants (Pauling et al., 2001). Comparative DNA sequencing analyses versus the wild-type identified all six strains as carrying distal nifE insertions (results not presented).


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

 
To measure dinitrogenase activity by acetylene reduction assay, the six nifE mutants were batch-cultured in liquid defined salts medium (ORS-MM) containing 10 mM ammonium at 30 °C to early-exponential phase, pelleted, washed and physiologically shifted to N-limited (NIF) medium (Donald et al., 1985) under constant sparging with O2/CO2/N2 (1 % : 1 % : 98 %). The N2 fixation-induced cultures were then divided; subcultures were incubated at 30 or 42 °C and at various times culture samples were removed for dinitrogenase activity assays as measured by acetylene reduction (see Methods). Among the nifE null strains, absolute rates for acetylene reduction activities at the permissive temperature (30 °C) were approximately 20 % those of wild-type strain 57100. And, in all six cases, residual, relative, fractional acetylene reduction activities (activity at 42 °C÷activity at 30 °C) were similar to that of the wild-type. By contrast, control nifD228 null strain 60228 showed no detectable acetylene reduction activity at either the restrictive or the permissive temperature. Therefore, when assayed by acetylene reduction, dinitrogenase activities of nifE insertion mutants did not show relative thermolability. Thus, from the insertion mutant screen, novel A. caulinodans gene(s) specifically required for high-temperature (42 °C) N2 fixation were not identified. As this insertion mutant search was exhaustive, A. caulinodans seemingly lacks genes specifically required for high-temperature (42 °C) N2 fixation.

Isolation of A. caulinodans N2 fixation temperature-sensitive (Nifts) thermolabile point mutants
Taking a more subtle approach, in a second genetic screen, we sought Nifts thermolabile point mutants arising from ethyl methanesulfonate mutagenesis of A. caulinodans 57100 (see Methods). Ethyl methanesulfonate-mutagenized cultures were outgrown in rich medium, shifted to N2 fixation-dependent liquid growth conditions at the restrictive temperature (42 °C) and subjected to a penicillin-type enrichment scheme (see Methods). After four growth/selection cycles, enriched cells were duplicate-plated and tested for specific, defective N2-dependent growth at 42 °C as discussed above. Candidates were retested for wild-type N2-dependent growth at 30 °C; identified strains were presumed to carry novel Nifts alleles. To help categorize new Nifts mutations, candidates were subjected to genetic complementation tests. To each candidate, recombinant plasmids pNSN8, pNSN11, pMSM15, pNSN18, pNSN21, pNSN31 and pNSN41 (Table 1) were introduced by electroporation; these recombinant plasmids carry previously identified A. caulinodans wild-type nif loci (Donald et al., 1986). Each of the 37 newly identified Nifts mutants was complemented by a particular pNSN plasmid at the restrictive temperature (results not presented). Therefore, no novel nif gene loci were uncovered.

Two Nifts mutants, strains 62106 and 62111 (Table 1), were complemented to wild-type by pNSN21, which carries wild-type N2 fixation locus-2, including the fixABCX operon (Donald et al., 1986). The fixAB product encodes ETFN; A. caulinodans recombinant ETFN has been purified and shown to comprise an FAD-containing heterodimer (results not presented). Residues of both ETFN subunits show approximately 40 % identity and 60 % conservation when compared with canonical ETF, both bacterial and eukaryotic. Within the Rhizobiaceae, the ETFN family is even more highly conserved; residues of both subunits typically show 70–80 % identity and 90 % conservation (Arigoni et al., 1991; Weidenhaupt et al., 1996). ETF mediates electrochemical coupling of multiple flavoprotein dehydrogenases (McKean et al., 1983). In the Rhizobiaceae, while ETFN functions specifically in N2 fixation (Donald et al., 1986; Arigoni et al., 1991), the presumed flavoprotein dehydrogenases with which ETFN electrochemically interacts are not known.

To refine map positions for these Nifts mutations, strains 62106 and 62111 were subjected to recombination tests. Plasmids pVi60101 (FixA+B+) and pVi60113 (FixB+C+X+) were introduced into strains 62106 and 62111 by electroporation, and cells were plated onto NIF medium at 42 °C. Thermostable Nif+ recombinants were obtained in the presence of both plasmids for both mutants, implying that strains 62106 and 62111 both carried fixB mutations. Accordingly, the fixB gene sequences of both strains were determined (see Methods) and compared to that of the wild-type (Arigoni et al., 1991). Both fixB temperature-sensitive alleles encoded distinct single-codon substitutions; the fixB238 missense allele of strain 62106 comprised an ETFN{alpha} Y238H substitution, whereas the fixB229 missense allele of strain 62111 comprised an ETFN{alpha} D229G substitution (Table 1). In the rhizobial ETFN{alpha} family, residues Y238 and D229 are highly conserved. Presumably, both missense alleles yield thermolabile ETFN variants, which fold aberrantly and are possibly impaired in flavoprotein–flavoprotein interactions (see Discussion).

Extragenic suppressors of A. caulinodans fixB temperature-sensitive mutants map to lpdA
Spontaneous, thermostable derivatives of strains 62106 and 62111 were selected after four successive subcultures in liquid NIF medium at 42 °C. Subcultures were plated onto solid NIF medium at 42 °C and wild-type, opaque colonies were identified, isolated and re-tested for both colony morphology on the same medium and diazotrophic growth in liquid culture at the restrictive temperature (see Methods). To analyse the fixB alleles of each thermostable derivative, primer pairs diagnostic both for wild-type and mutant nucleotide sequences were used in PCR amplification experiments (see Methods). Using 17 nt oligodeoxynucleotide primers specific for both fixB temperature-sensitive alleles, as well as a primer pair representing wild-type fixB, each candidate strain yielded a single, diagnostic PCR amplification product corresponding to either the wild-type or a parental fixB temperature-sensitive missense allele (see Methods). From these diagnostic PCR results, nine of 12 independent, thermostable derivatives of strain 62106 were assigned as true wild-type revertants; three of the 12 derivatives retained the parental fixB mutant allele and thus carried some extragenic suppressor mutation(s). For strain 62111, seven of its 12 derivatives were wild-type revertants, and five of the 12 carried extragenic suppressors. In repeated growth tests with NIF medium in both liquid batch cultures and on solid media, all suppressor strains appeared fully wild-type at the restrictive temperature (42 °C).

While initial hypotheses about the nature of such suppressor strains were not supported by experimentation, we were aided in the analysis of these strains by a serendipitous discovery. In unrelated work, we had analysed thermolabile derivatives of A. caulinodans strain 62004 that lacked PDH complex activity (Pauling et al., 2001). Unwittingly, two laboratory staff had given identical strain numbers to, in one case, thermolabile 62004 derivatives and, in the other, thermostable revertant/suppressor 62106 and 62111 derivatives. By happenstance, the latter strains were cultured and analysed for PDH activity in crude cell extracts. All but two strains showed wild-type PDH activity, measured as pyruvate-dependent reduction of NAD+ (see Methods). After re-numbering the collection, these two strains were correctly labelled as thermostable 62106 and 62111 derivatives. Upon retesting, these fixB temperature-sensitive suppressor strains, renumbered 62202 and 62223 (Table 1), again showed approximately 60 and 70 % of the wild-type PDH activity, respectively (Table 2). As evidenced by diagnostic PCR amplification and DNA sequencing, strains 62202 and 62223 retained a parental fixB temperature-sensitive missense allele and thus carried extragenic suppressor mutation(s).


View this table:
[in this window]
[in a new window]
 
Table 2. PDH activities in A. caulinodans strains

 
Crude cell extracts of strains 62202 and 62223 were also tested for membrane respiration-linked pyruvate oxidation activity, measured as pyruvate-dependent reduction of cytochrome c, coupled to artificial electron acceptors (see Methods). As measured by cytochrome c-dependent assay, pyruvate oxidation activity was present at essentially wild-type levels in both cases (Table 2). Therefore, PDH activity in strains 62202 and 62223 was partially defective.

To strengthen this inference, fixB temperature-sensitive parental strains 62106 and 62111 were dosed with extra copies of the wild-type pdh operon and tested for thermostable N2 fixation. Plasmid pTROY13 (Ludwig, 1987), which carries a functional, recombinant E. coli lamB gene, was introduced into fixB temperature-sensitive strains 62106 and 62111 by intergeneric bacterial conjugations, using E. coli SM10 as donor (Table 1). Because this lamB gene encodes a trimeric porin, localized to outer membranes, and which mediates maltodextrin transport, and because the LAMB porin gratuitously serves as a receptor for coliphage {lambda} (Randall-Hazelbauer & Schwartz, 1973; Schwartz, 1975), diverse Gram-negative bacteria expressing/exporting a recombinant LAMB are efficiently infected by DNA packaged as coliphage {lambda} (Ludwig, 1987). Liquid batch cultures of A. caulinodans strains 62106(pTROY13) and 62111(pTROY13) were infected at a high m.o.i. (~100 cell-1) with recombinant {lambda}pdhC4 carrying the wild-type A. caulinodans pdh operon (Pauling et al., 2001), and N2 fixation activity was induced by physiological shift (see Methods). When tested 4 h post-infection at the restrictive temperature (42 °C), {lambda}pdhC4-infected strains 62106(pTROY13) and 62111(pTROY13) reproducibly showed increased acetylene reduction activity, albeit not at fully wild-type level; in the absence of added recombinant {lambda}pdhC4, no increases in acetylene reduction activity at 42 °C were noted (results not presented). Therefore, both fixB temperature-sensitive strains were partially suppressed by obtaining additional copies of the wild-type pdh operon; phenotypic changes were presumably gene dosage effects. Moreover, this phenotypic change was similar to that obtained with thermostable, extragenic suppressors mapping to the pdh operon.

A. caulinodans strain 62004R, which carries a null pdhB allele and is absolutely defective for PDH complex activity, shows pronounced growth and N2 fixation phenotypes when tested in culture (Pauling et al., 2001). Accordingly, suppressor strains 62202 and 62223 were subjected to various physiological tests in liquid batch cultures at 30 °C for comparison to both Pdh- control strain 62004R and wild-type strain 57100. For the suppressor strains, growth phenotypes and N2 fixation activities in aerobic and microaerobic conditions showed either very subtle or no significant differences. Likewise, while both strains carried fixB temperature-sensitive mutations, no significant growth differences at either the permissive or the restrictive temperature were observed.

To characterize extragenic suppression precisely, pdh operon nucleotide sequences from thermostable suppressor strains 62202 and 62223 were determined (see Methods). In both suppressor strains, the pdhA, pdhB and pdhC genes proved identical to those of wild-type strain 57100. Therefore, nucleotide sequencing was extended distal to the pdhABC genes. Several hundred nucleotides downstream of pdhC, a 1·4 kbp ORF (GenBank accession no. AY331184) was identified. From multiple alignments of the translated product, this ORF was conclusively identified as that encoding LpDH, the third (E3) component of the PDH complex. For comparison, nucleotide sequences for the lpdA gene in suppressor strains 62202 and 62223 were also obtained (see Methods). For strains 62202 and 62223, single point mutations in the lpdA coding sequence were identified and confirmed by double-stranded nucleotide sequencing analyses. When parental thermolabile strains 62106 and 62111 were similarly analysed both strains carried the wild-type lpdA allele.

The distinctive lpdA point mutations carried by suppressor strains 62202 and 62223 resulted in single-codon substitutions of the inferred translated product LpDH, 472 residues in length. In strain 62202, the LpDH product carried a G187S substitution, whereas in strain 62223, LpDH carried an E210G substitution (Table 1). Among the {alpha}-Proteobacteria, the LpDH gene family is highly conserved. Indeed, atomic resolution crystal structures for both Azotobacter vinelandii LpDH (PDH complex) and Pseudomonas putida LpDH (2-oxoisovalerate dehydrogenase) are virtually superimposable (Mattevi et al., 1991, 1992). Accordingly, when the Azorhizobium caulinodans LpdA sequence was superimposed on the P. putida LpDH structure, residues 187 and 210 both mapped to the NAD+-binding domain and, as a result, the missense suppressor variants quite possibly adversely affected NAD+ binding.

To test this possibility, active PDH complex was purified from mid-exponential phase, aerobic batch cultures of wild-type, 62202 and 62223 grown in defined medium with L-malate as the energy source (see Methods). Values of Km (apparent) for NAD+ binding were measured for each PDH complex preparation under standard assay conditions (see Methods). Wild-type PDH complex yielded an apparent Km (NAD) value of 0·025 mM; 62202 PDH complex yielded an apparent Km (NAD) value of 0·13 mM; 62223 PDH complex yielded an apparent Km (NAD) value of 0·10 mM. Thus, the PDH complex activities comprising LpDH variants G187S and E210G showed four- to fivefold diminished NAD+-binding affinities. Moreover, wild-type PDH activity was strongly inhibited by the addition of 10 µM NADH. Such inhibition is likely to be allosteric. Therefore, in microaerobic wild-type A. caulinodans cultures, in which steady-state NADH-to-NAD+ ratios approach unity (Pauling et al., 2001), in vivo PDH activity might then be NADH-inhibited.

In summary, distinct lpdA alleles were found to suppress distinct fixB thermolabile alleles, allowing N2 fixation. As this extragenic suppressor analysis yielded a gene-for-gene interaction between A. caulinodans fixB and lpdA, physical interaction of the respective gene products ETFN and LpDH is implicit. Presumably the ETFN–LpDH physical interaction mediates electron transfer between these two flavoproteins. Thermolabile ETFN variants most likely possess decreased affinity for LpDH; suppressor LpDH variants compensate with decreased affinity for NAD+.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In both symbiotic N2-fixing and diazotrophic bacteria, contiguous fixAB genes encode heterodimeric ETFN. In A. caulinodans, single fixA or fixB null mutants are phenotypically Nif- in culture but are otherwise wild-type. Thus, ETFN is specifically required for dinitrogenase complex activity (Donald et al., 1986). In the Rhizobiaceae, fixAB reside in a highly conserved fixABCX operon (Corbin et al., 1983; Earl et al., 1987; Dusha et al., 1987; Arigoni et al., 1991). As assigned by gene order, fixA encodes ETFN{beta} and fixB encodes ETFN{alpha}. The ETF family shows the strongly conserved folded structures representative of the pyridine-nucleotide disulfide oxidoreductase superfamily. In bacteria and eukaryotes, the ETF heterodimer tightly but non-covalently binds a single FAD and a single 5'-AMP, the latter an evolutionary vestige of the conserved pyridine-nucleotide site typical of this superfamily. However, the ETF family lacks both an active disulfide and its participating cysteine residues otherwise conserved among this superfamily. In electrochemical studies, ETF shows both one- and/or two-electron transfer to/from various flavoprotein dehydrogenases. In eukaryotic mitochondria, principal electron donors to ETF include the acyl-CoA dehydrogenases of fatty-acid catabolism. However, as the Rhizobiaceae do not actively catabolize aliphatic fatty acids for use as an energy source, ETFN must serve as an electron acceptor to other flavoprotein dehydrogenases.

In eukaryotic mitochondria, a peripheral, membrane-associated ETF : ubiquinone oxidoreductase (ETF-QO) serves as the principal electron acceptor to ETF, which then couples mitochondrial oxidations by soluble (matrix) flavoprotein dehydrogenases to membrane respiration (Goodman et al., 1994). However, in the Rhizobiaceae, ETFN-dependent membrane respiration characteristic of ETF-QO activity has not been observed. From primary sequence homology, the FIXC family is a highly conserved flavoprotein distantly related to ETF-QO yet lacking both hydrophobic and amphipathic transmembrane {alpha}-helices. Thus, the FIXC family is unlikely to be stably membrane-associated. In the Rhizobiaceae, the tightly linked fixX gene encodes a typical, small, soluble [4Fe–4S]ferredoxin (Arigoni et al., 1991). In diverse microaerophilic N2-fixing bacteria, because the fixABCX gene set is both highly conserved and invariably tightly linked, FIXC likely comprises an ETF : ferredoxin oxidoreductase. If so, then a plausible (soluble) electron transfer pathway is

What flavoprotein(s) serve as the electron donor to ETFN? In the Rhizobiaceae, the trail blazed by genetic analysis of N2 fixation abruptly ends; despite saturation mutagenesis experiments, no Nif- mutant representing a putative electron donor to ETFN has been identified. This result leaves two possible explanations, which are not mutually exclusive: (i) redundancy and (ii) pleiotropy. In support of the former, if ETFN functions to recruit multiple electron donors for N2 fixation, any single genetic defect in a particular ETFN electron donor might not necessarily render a phenotype. In support of the latter, there might exist a specific ETFN electron donor in which mutations would yield multiple growth defects and not a simple Nif- phenotype.

Consistent with pleiotropy, we have here described a gene-for-gene interaction between the A. caulinodans fixB and lpdA genes. This genetic interaction implies physical interaction between the respective products ETFN{alpha} and the LpDH paralogue comprising the E3 component of the PDH complex. This physical interaction presumably facilitates electron transfer. If the PDH complex functions as electron donor to ETFN, then pyruvate oxidation is the ultimate metabolic source of reducing equivalents for dinitrogenase complex activity (Fig. 1). Indeed, A. caulinodans strain 62004 carrying the null pdhB4 mutation shows a conditional Nif- phenotype. However, the PDH complex also serves as the driving force for oxidative phosphorylation as generally required for N2 fixation (Pauling et al., 2001). Might the PDH complex serve N2 fixation in both capacities?



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 1. A. caulinodans electron-transfer pathways in microaerobic culture in support of N2 fixation. NAD+ and ETFN kinetically compete to re-oxidize the PDH complex. Membrane electron (e-) transport (oxidative phosphorylation) yields ATP; soluble electron transport to N2 consumes ATP.

 
In aerobic culture, the PDH complex LpDH (E3) component uses NAD+ as an electron acceptor. During microaerobic N2 fixation, both NAD+ and ETFN might kinetically compete to re-oxidize the FADH2 centre of LpDH. Is this kinetic competition plausible? In the Rhizobiaceae, N2 fixation is optimally adapted to true microaerobic conditions. When shifted from fully aerobic conditions, A. caulinodans microaerobic cultures show 10-fold increases in steady-state in vivo NADH-to-NAD+ ratios, approaching unity (Pauling et al., 2001). The archetype E. coli LpDH activity, a freely reversible enzyme, shows quite similar standard mid-point potentials for both the dihydrolipoamide and NADH two-electron electrochemical half-cell reactions (Koike et al., 1960). From equilibrium considerations, in vivo LpDH activity should theoretically tend to exhibit less regulation as the intracellular NADH-to-NAD+ ratio approaches unity. However, NADH is itself a powerful allosteric inhibitor of E. coli LpDH activity (Koike et al., 1960). Similarly, added 10 µM NADH yields strong inhibition of purified A. caulinodans PDH activity. Thus, in A. caulinodans microaerobic cultures, LpDH activity might well be allosterically NADH-inhibited, possibly affecting its response to either NAD+ or ETFN as oxidant. As the fixABCX operon encoding ETFN is derepressed by FIXK, itself specifically produced in response to microaerobic culture (Kaminski et al., 1991), ETFN would necessarily then be more abundant.

From comparisons of stably folded structures, several inferences may be drawn from the thermolabile ETFN mutants and their suppressor LpDH mutants. In Paracoccus denitrificans, the ‘housekeeping’ ETF has a folded structure highly conserved with that of mammalian ETF (Herrick et al., 1994). ETF{alpha} folds into two globular domains, one of which binds FAD, the second acts to form a flexible hinge, stabilizing flavoprotein–flavoprotein interactions; ETF{beta} folds into a single domain, which binds 5'-AMP. If rhizobial ETFN folds similarly, then the thermolabile mutations ETFN{alpha} Y238H (fixB238) and ETFN{alpha} D229G (fixB229) both map to the ETFN{alpha} interdomain loop, in which single residue substitutions would be unlikely per se to affect stable ETFN folding. Rather, the observed thermolability of these ETFN variants might be owed to weakened flavoprotein–flavoprotein interactions, presuming the ETFN{alpha} interdomain loop were to help stabilize such interactions.

What of the suppressor LpDH single residue variants? From genome analyses, {alpha}-Proteobacteria typically have three or more dispersed, paralogous lpd genes encoding LpDH isoforms. Discrete lpd genes map to loci encoding pyruvate, 2-oxoglutarate and 2-oxoisovalerate dehydrogenase complex activities. (In the {alpha}-Proteobacteria, an unlinked, fourth lpd gene typically encodes an exported, periplasmic LpDH activity replete with its own lipoamide-anchor domain.) These three oxoacid dehydrogenase complexes share a conserved reaction mechanism. Atomic resolution crystal structures for both Azotobacter vinelandii LpDH (PDH complex) and P. putida LpDH (2-oxoisovalerate dehydrogenase complex) are virtually superimposable (Mattevi et al., 1991, 1992). Rhizobiaceae LpDH isoforms show strong primary sequence homology with both P. putida LpDH (PDH) and A. vinelandii LpDH (2-oxoisovalerate dehydrogenase). Indeed, the atomic resolution crystal structure of P. putida LpDH complexed with substrate NAD+ is particularly informative (Mattevi et al., 1992). Recall that in Azorhizobium caulinodans, the extragenic fixB temperature-sensitive suppressors include both lpdA187 and lpdA210 alleles which yield LpDH single-residue substitutions G187S and E210G. Both substitutions affect conserved residues of the NAD+-binding domain, as defined for Azotobacter vinelandii LpDH. Indeed, as evidenced here, these substitutions yield variant LpDH products with four- to fivefold decreased NAD+ affinities.

An obvious model for the lpdAfixB gene-for-gene interaction follows from these inferences: ETFN{alpha} Y238H and ETFN{alpha} D229G both suffer decreased ETFN–LpDH affinity; LpDH G187S and LpDH E210G both show decreased NAD+–LpDH affinity. The phenotypic effects of these decreased affinities are thus compensatory.

Other flavoprotein dehydrogenases might conceivably serve as electron donors to ETFN. Parsimoniously, our results simply correlate certain ETFN loss-of-function alleles with certain LpDH (PDH) gain-of-function alleles. (In formal genetic parlance, ‘gain-of-function’ – gain of property ‘B’ – may fortuitously result from loss of property ‘A’.) Given ETF's general role of facilitating the oxidation–reduction balance among multiple flavoprotein dehydrogenases, and given that Azorhizobium caulinodans possesses multiple (including PDH, 2-oxoglutarate dehydrogenase and 2-oxoisovalerate dehydrogenase) flavoprotein dehydrogenase complexes, might not other flavoproteins donate electrons to ETFN? The rhizobial LpDH (2-oxoglutarate dehydrogenase) gene family carries an additional anchor domain conserved among peripheral membrane proteins. Quite possibly, this domain links rhizobial 2-oxoglutarate dehydrogenase complex activity directly to membrane-bound quinones as electron acceptor, precluding any interaction with ETFN. Regarding the 2-oxoisovalerate dehydrogenase complex, inferences are problematic. While A. caulinodans avidly catabolizes the branched-chain amino acids leucine, isoleucine and valine, the use of these amino acids as defined energy source results in steady-state nitrogen excess, a physiological condition that represses N2 fixation. In A. caulinodans, cognate 2-oxoacids (products of aminotransferase activities) of leucine, isoleucine and valine are taken up very poorly and sustain little or no culture growth.

In an analogous fashion, Klebsiella pneumoniae, a microaerobic diazotroph unrelated to the Rhizobiaceae, employs a specific flavodoxin-dependent pyruvate oxidoreductase activity as a metabolic electron donor for N2 fixation (Hill & Kavanagh, 1980). Notably, this pyruvate : flavodoxin oxidoreductase is completely absent from the Rhizobiaceae.

In collaborative experiments, A. caulinodans recombinant ETFN has been purified and its active FAD centre has been isolated. However, electrochemical properties of purified ETFN, including operative mid-point oxidation–reduction potentials, are markedly perturbed by the addition of other {alpha}-proteobacterial pyridine-nucleotide disulfide oxidoreductase flavoproteins, rendering any conclusions problematic. Quite possibly, given the ‘flexible-hinge’ folding model, flavoprotein–flavoprotein complex formation might precipitate large conformational changes in ETF, substantially altering its operative electrochemical oxidation–reduction potential (Byron et al., 1989; Jones et al., 2000).

From purely kinetic considerations, ETF's participation in soluble electron-transfer pathways would seem of dubious merit. As higher-order complex formation lacks both theoretical and experimental support, ETF presumably forms only binary complexes with other flavoprotein dehydrogenases, alternately serving as electron acceptor and electron donor. If so, in vivo electron-transfer rates would reflect successive, soluble flavoprotein–flavoprotein formation/dissociation steps, and apparent rate constants would then be higher-order with respect to ETF concentrations. Among (evolutionarily) modern organisms, and given these kinetic constraints, N2 fixation, with its exceedingly low kcat value (~10 s-1), is perhaps one of the very few metabolic processes accepting of ETFN as an obligate electron-transfer intermediary.

What of the unusual capacity of A. caulinodans to carry out N2 fixation at a relatively high temperature (42 °C)? In various diazotrophic bacteria, transcriptional regulation in general, and the NifA transcriptional activator in particular, seem to exhibit thermolability (Zhu & Brill, 1981). In the present study with A. caulinodans, thermolabile nifA mutants were absent. Rather, thermolabile nifE null mutants were obtained. Yet, while nifE null mutants have been identified in various diazotrophic bacteria (MacNeil et al., 1978; Evans et al., 1988), a thermolabile phenotype was not noted. NifE is purported to be important for biogenesis of the iron–molybdenum cofactor, itself required for dinitrogenase complex activity (Roberts et al., 1978; Imperial et al., 1987). Indeed, in the present study, A. caulinodans null mutants showing absolute thermolability for N2 fixation were not identified. While such ‘negative results' must be cautiously interpreted, as multiple, independent nifE mutants with a partially defective phenotype were indeed isolated, it is tempting to regard this search as exhaustive. If so, then A. caulinodans lacks particular gene(s) specifically required for high temperature N2 fixation and likely possesses an orthodox nif gene set whose alleles are collectively adapted to relatively thermostable N2 fixation.


   ACKNOWLEDGEMENTS
 
This work was supported by grants from the US National Science Foundation and the US National Institutes of Health to R. A. L.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Arigoni, F., Kaminski, P. A., Hennecke, H. & Elmerich, C. (1991). Nucleotide sequence of the fixABC region of Azorhizobium caulinodans ORS571: similarity of the fixB product with eukaryotic flavoproteins, characterization of fixX, and identification of nifW. Mol Gen Genet 225, 514–520.[Medline]

Byron, C. M., Stankovich, M. T., Husain, M. & Davidson, V. L. (1989). Unusual redox properties of electron-transfer flavoprotein from Methylophilus methylotrophus. Biochemistry 28, 8582–8587.[Medline]

Corbin, D., Barran, L. & Ditta, G. (1983). Organization and expression of Rhizobium meliloti nitrogen fixation genes. Proc Natl Acad Sci U S A 80, 3005–3009.[Abstract]

Dallo, S. F., Kannan, T. R., Blaylock, M. W. & Baseman, J. B. (2002). Elongation factor Tu and E1 beta subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol Microbiol 46, 1041–1051.[CrossRef][Medline]

Donald, R. G. K., Raymond, C. K. & Ludwig, R. A. (1985). Vector insertion mutagenesis of Rhizobium sp. strain ORS571: direct cloning of mutagenised DNA sequences. J Bacteriol 162, 317–323.[Medline]

Donald, R. G. K., Nees, D., Raymond, C. K., Loroch, A. I. & Ludwig, R. A. (1986). Three genomic loci encode Rhizobium sp. ORS571 N2 fixation genes. J Bacteriol 165, 72–81.[Medline]

Dusha, I., Kovalenko, S., Banfalvi, Z. & Kondorosi, A. (1987). Rhizobium meliloti insertion element ISRm2 and its use for identification of the fixX gene. J Bacteriol 169, 1403–1409.[Medline]

Earl, C. D., Ronson, C. W. & Ausubel, F. M. (1987). Genetic and structural analysis of the Rhizobium meliloti fixA, fixB, fixC, and fixX genes. J Bacteriol 169, 1127–1136.[Medline]

Evans, D., Jones, R., Woodley, P. & Robson, R. (1988). Further analysis of nitrogen fixation (nif) genes in Azotobacter chroococcum: identification and expression in Klebsiella pneumoniae of nifS, nifV, nifM and nifB genes and localization of nifE/N-, nifU-, nifA- and fixABC-like genes. J Gen Microbiol 134, 931–942.[Medline]

Finocchiaro, G., Ito, M., Ikeda, Y. & Tanaka, K. (1988). Molecular cloning and nucleotide sequence of cDNAs encoding the alpha-subunit of human electron transfer flavoprotein. J Biol Chem 263, 15773–15780.[Abstract/Free Full Text]

Goodman, S. I., Axtell, K. M., Bindoff, L. A., Beard, S. E., Gill, R. E. & Frerman, F. E. (1994). Molecular cloning and expression of a cDNA encoding human electron transfer flavoprotein-ubiquinone oxidoreductase. Eur J Biochem 219, 277–286.[Abstract]

Herrick, K. R., Salazar, D., Goodman, S. I., Finocchiaro, G., Bedzyk, L. A. & Frerman, F. E. (1994). Expression and characterization of human and chimeric human Paracoccus denitrificans electron transfer flavoprotein. J Biol Chem 269, 32239–32245.[Abstract/Free Full Text]

Hill, S. & Kavanagh, E. P. (1980). Roles of nifF and nifJ gene products in electron transport to nitrogenase in Klebsiella pneumoniae. J Bacteriol 141, 470–475.[Medline]

Imperial, J., Shah, V. K., Ugalde, R. A., Ludden, P. W. & Brill, W. J. (1987). Iron-molybdenum cofactor synthesis in Azotobacter vinelandii Nif- mutants. J Bacteriol 169, 1784–1786.[Medline]

Jones, M., Basran, J., Sutcliffe, M. J., Gunter-Grossmann, J. & Scrutton, N. S. (2000). X-ray scattering studies of Methylophilus methylotrophus (sp. W3A1) electron-transferring flavoprotein. Evidence for multiple conformational states and an induced fit mechanism for assembly with trimethylamine dehydrogenase. J Biol Chem 275, 21349–21354.[Abstract/Free Full Text]

Kaminski, P. A., Mandon, K., Arigoni, F., Desnoues, N. & Elmerich, C. (1991). Regulation of nitrogen fixation in Azorhizobium caulinodans: identification of a fixK-like gene, a positive regulator of nifA. Mol Microbiol 5, 1983–1991.[Medline]

Koike, M., Shah, P. C. & Reed, L. J. (1960). {alpha}-Ketoacid dehydrogenation complexes. III. Purification and properties of dihydrolipoic dehydrogenase of Escherichia coli. J Biol Chem 235, 1939–1943.[Medline]

Kwon, D. K. & Beevers, H. (1992). Growth of Sesbania rostrata (brem) with stem nodules under controlled conditions. Plant Cell Environ 15, 939–945.

Ludwig, R. A. (1987). Gene tandem-mediated selection of coliphage {lambda}-receptive Agrobacterium, Pseudomonas, and Rhizobium strains. Proc Natl Acad Sci U S A 84, 3334–3338.[Abstract]

MacNeil, T., MacNeil, D., Roberts, G. P. & Brill, W. J. (1978). Fine-structure mapping and complementation analysis of nif (nitrogen fixation) genes in Klebsiella pneumoniae. J Bacteriol 136, 253–266.[Medline]

Mattevi, A., Schierbeek, A. J. & Hol, W. G. (1991). Refined crystal structure of lipoamide dehydrogenase from Azotobacter vinelandii at 2·2 Å resolution: a comparison with the structure of glutathione reductase. J Mol Biol 220, 975–994.[Medline]

Mattevi, A., Obmolova, G., Sokatch, J. R., Betzel, C. & Hol, W. G. (1992). The refined crystal structure of Pseudomonas putida lipoamide dehydrogenase complexed with NAD+ at 2·45 Å resolution. Proteins 13, 336–351.[Medline]

McKean, M. C., Beckman, J. D. & Frerman, F. E. (1983). Subunit structure of electron transfer flavoprotein. J Biol Chem 258, 1866–1870.[Abstract/Free Full Text]

O'Neill, H., Mayhew, S. G. & Butler, G. (1998). Cloning and analysis of the genes for a novel electron-transferring flavoprotein from Megasphaera elsdenii. J Biol Chem 273, 21015–21024.[Abstract/Free Full Text]

Pauling, D. C., Lapointe, J. P., Paris, C. M. & Ludwig, R. A. (2001). Azorhizobium caulinodans pyruvate dehydrogenase activity is dispensable for aerobic but required for microaerobic growth. Microbiology 147, 2233–2245.[Abstract/Free Full Text]

Pühler, A., Burkhardt, H. J. & Klipp, W. (1979). Cloning of the entire region for nitrogen fixation from Klebsiella pneumoniae on a multicopy plasmid vehicle in Escherichia coli. Mol Gen Genet 176, 17–24.[Medline]

Randall-Hazelbauer, L. & Schwartz, M. (1973). Isolation of the bacteriophage lambda receptor from Escherichia coli. J Bacteriol 116, 1436–1446.[Medline]

Roberts, G. P., MacNeil, T., MacNeil, D. & Brill, W. J. (1978). Regulation and characterization of protein products coded by the nif (nitrogen fixation) genes of Klebsiella pneumoniae. J Bacteriol 136, 267–279.[Medline]

Roberts, D. L., Salazar, D., Fulmer, J. P., Frerman, F. E. & Kim, J.-J. P. (1999). Crystal structure of Paracoccus denitrificans electron transfer flavoprotein: structural and electrostatic analysis of a conserved flavin-binding domain. Biochemistry 38, 1977–1989.[CrossRef][Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schwartz, M. (1975). Reversible interaction between coliphage lambda and its receptor protein. J Mol Biol 99, 185–201.[Medline]

Simon, R., Priefer, U. & 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.

Smith, R. F. & Smith, T. F. (1990). Automatic generation of primary sequence patterns from sets of related protein sequences. Proc Natl Acad Sci U S A 87, 118–122.[Abstract]

Weidenhaupt, M., Rossi, P., Beck, C., Fischer, H.-M. & Hennecke, H. (1996). Bradyrhizobium japonicum possesses two discrete sets of electron transfer flavoprotein genes: fixA, fixB and etfS, etfL. Arch Microbiol 165, 169–178.[CrossRef][Medline]

Zhu, J. & Brill, W. J. (1981). Temperature sensitivity of the regulation of nitrogenase synthesis by Klebsiella pneumoniae. J Bacteriol 145, 1116–1118.[Medline]

Received 27 June 2003; revised 17 September 2003; accepted 1 October 2003.



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 Scott, J. D.
Articles by Ludwig, R. A.
Articles citing this Article
PubMed
PubMed Citation
Articles by Scott, J. D.
Articles by Ludwig, R. A.
Agricola
Articles by Scott, J. D.
Articles by Ludwig, R. A.


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.