Isolation and characterization of nitrogenase-derepressed mutant strains of cyanobacterium Anabaena variabilis.

A positive selection method for isolation of nitrogenase-derepressed mutant strains of a filamentous cyanobacterium, Anabaena variabilis, is described. Mutant strains that are resistant to a glutamate analog, L-methionine-D,L-sulfoximine, were screened for their ability to produce and excrete NH4+ into medium. Mutant strains capable of producing nitrogenase in the presence of NH4+ were selected from a population of NH4+-excreting mutants. One of the mutant strains (SA-1) studied in detail was found to be a conditional glutamine auxotroph requiring glutamine for growth in media containing N2, NO3-, or low concentrations of NH4+ (less than 0.5 mM). This glutamine requirement is a consequence of a block in the assimilation of NH4+ produced by an enzyme system like nitrogenase. Glutamate and aspartate failed to substitute for glutamine because of a defect in the transport and utilization of these amino acids. Strain SA-1 assimilated NH4+ when the concentration in the medium reached about 0.5 mM, and under these conditions the growth rate was similar to that of the parent. Mutant strain SA-1 produced L-methionine-D,L-sulfoximine-resistant glutamine synthetase activity. Kinetic properties of the enzyme from the parent and mutant were similar. Mutant strain SA-1 can potentially serve as a source of fertilizer nitrogen to support growth of crop plants, since the NH4+ produced by nitrogenase, utilizing sunlight and water as sources of energy and reductant, respectively, is excreted into the environment.     

Comparison of hup trait and intrinsic antibiotic resistance for assessing rhizobial competitiveness axenically and in soil

A defined medium for growth of 24 strains of Moraxella (Branhamella) catarrhalis was devised. This medium (medium B4) contains sodium lactate as a partial carbon source, proline as both a partial carbon source and a partial nitrogen source, aspartate as a partial nitrogen source, and the growth factors arginine, glycine, and methionine. Either aspartate, glutamate, or proline could serve as sole nitrogen source, but growth occurred at a significantly better rate if proline was present together with either aspartate or glutamate, or with both aspartate and glutamate. With the exception of strain ATCC 23246, all the strains had an absolute requirement for arginine and either a partial or absolute requirement for glycine. The concentration of glycine required for optimal growth was found to be relatively high for an amino acid growth factor. Heart infusion broth was found to be growth inhibitory for spontaneous mutants of one strain able to grow in the absence of arginine, and such mutants reverted readily to arginine dependence accompanied by the ability to grow faster on the complex medium. Growth rates in the defined medium B4 were enhanced by the simultaneous addition of asparagine, glutamate, glutamine, leucine, lysine, histidine, and phenylalanine.     

Comparison of Hup Trait and Intrinsic Antibiotic Resistance for Assessing Rhizobial Competitiveness Axenically and in Soil

A defined medium for growth of 24 strains of Moraxella (Branhamella) catarrhalis was devised. This medium (medium B4) contains sodium lactate as a partial carbon source, proline as both a partial carbon source and a partial nitrogen source, aspartate as a partial nitrogen source, and the growth factors arginine, glycine, and methionine. Either aspartate, glutamate, or proline could serve as sole nitrogen source, but growth occurred at a significantly better rate if proline was present together with either aspartate or glutamate, or with both aspartate and glutamate. With the exception of strain ATCC 23246, all the strains had an absolute requirement for arginine and either a partial or absolute requirement for glycine. The concentration of glycine required for optimal growth was found to be relatively high for an amino acid growth factor. Heart infusion broth was found to be growth inhibitory for spontaneous mutants of one strain able to grow in the absence of arginine, and such mutants reverted readily to arginine dependence accompanied by the ability to grow faster on the complex medium. Growth rates in the defined medium B4 were enhanced by the simultaneous addition of asparagine, glutamate, glutamine, leucine, lysine, histidine, and phenylalanine.     

Amino acids as repressors of nitrogenase biosynthesis in Klebsiella pneumoniae.

Nitrogenase biosynthesis in Klebsiella pneumoniae including mutant strains, which produce nitrogenase in the presence of NH+4 (Shanmugam, K.T., Chan, Irene, and Morandi, C. (1975) Biochim. Biophys. Acta 408, 101--111) is repressed by a mixture of L-amino acids. Biochemical analysis shows that glutamine synthetase activity in strains SK-24, SK-28, and SK-29 is also repressed by amino acids, with no detectable effect on glutamate dehydrogenase. Among the various amino acids, L-glutamine in combination with L-aspartate was found to repress nitrogenase biosynthesis completely. In the presence of high concentrations of glutamine (1 mg/ml) even NH+4 repressed nitrogenase biosynthesis in the strains SK-27, SK-37, SK-55 and SK-56. Under these conditions, increased glutamate dehydrogenase activity was also detected. Physiological studies show that nitrogenase derepressed strains are unable to utilize NH+4 as sole source of nitrogen for biosynthesis of glutamate for biosynthesis of glutamate, whereas back mutations leading to NH+4 utilization results in sensitivity to repression by NH+4. These findings suggest that amino acids play an important role as regulators of nitrogen fixation.     

Potential roles for the glnB and ntrYX genes in Azospirillum brasilense

Three Azospirillum brasilense mutants constitutive for nitrogen fixation (Nif(C)) in the presence of NH4(+) and deficient in nitrate-dependent growth were used as tools to define the roles of the glnB and ntrYX genes in this organism. Mutant HM14 was complemented for nitrate-dependent growth and NH4(+) regulation of nitrogenase by plasmid pL46 which contains the ntrYX genes of A. brasilense. Mutant HM26 was restored for NH4(+) regulation and nitrate-dependent growth by plasmid pJC1, carrying the A. brasilense glnB gene expressed from a constitutive promoter. Mutant HM053, on the other hand, was not complemented for NH4(+) regulation of nitrogenase and nitrate-dependent growth by both plasmids pJCI and pL46. The levels and control of glutamine synthetase activity of all mutants were not affected by both plasmids pL46 (ntrYX) and pJC1 (glnB). These results support the characterization of strains HM14 as an ntrYX mutant and strain HM26 as a glnB mutant and the involvement of ntrYX and glnB in the regulation of the general nitrogen metabolism in A. brasilense.     

Evolution of a recombinant (gucoamylase-producing) strain of Fusarium venenatum A3/5 in chemostat culture

Fusarium venenatum JeRS 325 is a transformant of strain A3/5 which produces Aspergillus niger glucoamylase (GAM) under the control of a Fusarium oxysporum trypsin-like protease promoter. The evolution of JeRS 325 was studied in glucose-limited chemostat cultures grown on NaNO3 or (NH4)2SO4 as the nitrogen source. Thirteen mutants which were more highly branched and four mutants which were more sparsely branched than the parental strain were isolated from the NaNO3 chemostat. The highly branched mutants detected in this chemostat did not displace the sparsely branched population. The mutants isolated from the NaNO3 chemostat complemented representative strains previously isolated from glucose-limited chemostat cultures of F. venenatum A3/5 grown on (NH4)2SO4, but showed little complementation between themselves. By contrast, a highly branched mutant isolated from the (NH4)2SO4 chemostat culture displaced the sparsely branched mycelial population. None of the mutants isolated from the NaNO3 or (NH4)2SO4 chemostats produced as much GAM as JeRS 325. Southern blot analysis showed that all except one mutant had lost copies of both the glucoamylase and the acetamidase (the selectable marker) genes. However, specific GAM production was not necessarily correlated with the extent of glaA gene loss observed. Further, 10 of the mutants had lost the ability to grow on acetamide as the sole nitrogen source, although they retained copies of the amdS gene. In competition studies, mutants which could not utilize acetamide displaced mutants which could. The presence of foreign DNA in JeRS 325 resulted in a reduced specific growth rate (compared to A3/5), but the presence of the foreign DNA did not prevent the evolution of the strain or the isolation of mutants which had improved growth rates.     

Altered kinetic properties of tyrosine-183 to cysteine mutation in glutamine synthetase of anabaena variabilis strain SA1 is responsible for excretion of ammonium ion produced by nitrogenase

A L-methionine- D, L-sulfoximine-resistant mutant of the cyanobacterium Anabaena variabilis, strain SA1, excreted the ammonium ion generated from N(2) reduction. In order to determine the biochemical basis for the NH(4)(+)-excretion phenotype, glutamine synthetase (GS) was purified from both the parent strain SA0 and from the mutant. GS from strain SA0 (SA0-GS) had a pH optimum of 7.5, while the pH optimum for GS from strain SA1 (SA1-GS) was 6.8. SA1-GS required Mn(+2) for optimum activity, while SA0-GS was Mg(+2) dependent. SA0-GS had the following apparent K(m) values at pH 7.5: glutamate, 1.7 m M; NH(4)(+), 0.015 m M; ATP, 0.13 m M. The apparent K(m) for substrates was significantly higher for SA1-GS at its optimum pH (glutamate, 9.2 m M; NH(4)(+), 12.4 m M; ATP, 0.17 m M). The amino acids alanine, aspartate, cystine, glycine, and serine inhibited SA1-GS less severely than the SA0-GS. The nucleotide sequences of glnA (encoding glutamine synthetase) from strains SA0 and SA1 were identical except for a single nucleotide substitution that resulted in a Y183C mutation in SA1-GS. The kinetic properties of SA1-GS isolated from E. coli or Klebsiella oxytoca glnA mutants carrying the A. variabilis SA1 glnA gene were also similar to SA1-GS isolated from A. variabilis strain SA1. These results show that the NH(4)(+)-excretion phenotype of A. variabilis strain SA1 is a direct consequence of structural changes in SA1-GS induced by the Y183C mutation, which elevated the K(m) values for NH(4)(+) and glutamate, and thus limited the assimilation of NH(4)(+) generated by N(2) reduction. These properties and the altered divalent cation-mediated stability of A. variabilis SA1-GS demonstrate the importance of Y183 for NH(4)(+) binding and metal ion coordination.     

Mechanism of l-Glutamine Production by an l-Glutamine-Producing Mutant of Flavobacterium rigense

Properties of some enzymes involved in l-glutamine biosynthesis in an l-glutamine-producing mutant of Flavobacterium rigense were examined. Glutamate-oxaloacetate transaminase in the mutant was nearly at the same level as that in the parent strain and was the most active among the enzymes participating in glutamate biosynthesis from alpha-ketoglutarate. Glutamine synthetase formation in the mutant was enhanced by increasing the concentration of (NH(4))(2)-fumarate in the medium, but the activity of this enzyme in the parent strain was very low, and its formation was not influenced by the concentration of (NH(4))(2)-fumarate. Glutaminase formation by both strains was similar and was not influenced by the levels of (NH(4))(2)-fumarate. Glutaminase activity of the mutant was inhibited by ammonia and fumarate. Intracellular amino acids and extracellular free amino acids in the mutant were compared with those of the parent strain. It seems reasonable to conclude that l-glutamine leaks out specifically through the cell membrane of strain 703 and that this specific excretion of l-glutamine probably allows a continuous conversion of l-glutamate to l-glutamine inside the cell.     

Rhizobium meliloti glutamate synthase: cloning and initial characterization of the glt locus.

The genetic locus glt, encoding glutamate synthase from Rhizobium meliloti 1021, was selected from a pLAFR1 clone bank by complementation of the R. meliloti 41 Glt- mutant AK330. A fragment of cloned DNA complementing this mutant also served to complement the Escherichia coli glt null mutant PA340. Complementation studies using these mutants suggested that glutamate synthase expression requires two complementation groups present at this locus. Genomic Southern analysis using a probe of the R. meliloti 1021 glt region showed a close resemblance between R. meliloti 1021, 41, and 102f34 at glt, whereas R. meliloti 104A14 showed many differences in restriction fragment length polymorphism patterns at this locus. R. meliloti 102f34, but not the other strains, showed an additional region with sequence similarity to glt. Insertion alleles containing transposable kanamycin resistance elements were constructed and used to derive Glt- mutants of R. meliloti 1021 and 102f34. These mutants were unable to assimilate ammonia and were Nod+ Fix+ on alfalfa seedlings. The mutants also showed poor or no growth on nitrogen sources such as glutamate, aspartate, arginine, and histidine, which are utilized by the wild-type parental strains. Strains that remained auxotrophic but grew nearly as well as the wild type on these nitrogen sources were readily isolated from populations of glt insertion mutants, indicating that degradation of these amino acids is negatively regulated in R. meliloti as a result of disruptions of glt.     

Salmonella typhimurium LT-2 mutants with altered glutamine synthetase levels and amino acid uptake activities.

To determine whether Salmonella typhimurium has a nitrogen control response, we have examined the regulation of nitrogen utilization in two mutants with fivefold and threefold elevations in their glutamine synthetase activities. The mutants do not require glutamine for growth on glucose--ammonia medium but do have altered growth on other nitrogen sources. They grow better than an isogenic control on media containing arginine or asparate, but more slowly with proline or alanine as nitrogen sources. This unusual growth pattern is not due to altered regulation of the ammonia assimilatory enzymes, glutamate dehydrogenase and glutamate synthase, or to changes in the enzymes for aspartate degradation. However, transport for several amino acids may be affected. Measurement of amino acid uptake show that the mutants with high glutamine synthetase levels have increased rates for glutamine, arginine, aspartate, and lysine, but a decreased rate for proline. The relationship between glutamine synthetase levels and uptake was examined in two mutants with reduced, rather than increased, glutamine synthetase production. The uptake rates for glutamine and lysine were lower in these two glutamine auxotrophs than in the Gln+ controls. These results show a correlation between the glutamine synthetase levels and the uptake rates for several amino acids. In addition, the pleiotropic growth of the mutants with elevated glutamine synthetase activities suggests that a nitrogen control response exists for S. typhimurium and that it can be altered by mutations affecting glutamine synthetase regulation.     

pH-Conditional, Ammonia Assimilation-Deficient Mutants of Hydrogenomonas eutropha: Isolation and Growth Characteristics

Two mutants of the facultative autotroph Hydrogenomonas eutropha were isolated by using a modified penicillin selection method. The mutation involved was unusual in that its effect on cellular growth was conditional with regard to extracellular pH and the type of substrate employed. Growth of both mutants was abnormal under autotrophic conditions and during heterotrophic cultivation in the presence of organic substrates which lacked an amino group. Abnormal growth was characterized by linear growth rates which were low at pH 6.0 and moderate at pH 7.2. In contrast, growth of the mutants was normal on most amino acids. Those substrates yielding abnormal growth were oxidized at normal rates by the mutants, indicating the mutation did not impair their uptake or metabolism. The data suggest that the mutants are defective in their ability to assimilate inorganic nitrogen into organic forms, and this defect is strongly influenced by pH.     

Regulation of nitrogen fixation. Nitrogenase-derepressed mutants of Klebsiella pneumoniae.

1. A new procedure is described for selecting nitrogenase-derepressed mutants based on the method of Brenchley et al. (Brenchley, J.E., Prival, M.J. and Magasanik, B. (1973) J. Biol. Chem. 248, 6122-6128) for isolating histidase-constitutive mutants of a non-N2-fixing bacterium. 2. Nitrogenase levels of the new mutants in the presence of NH4+ were as high as 100% of the nitrogenase activity detected in the absence of NH4+. 3. Biochemical characterization of these nitrogen fixation (nif) derepressed mutants reveals that they fall into three classes. Three mutants (strains SK-24, 28 and 29), requiring glutamate for growth, synthesize nitrogenase and glutamine synthetase constitutively (in the presence of NH4+). A second class of mutants (strains SK-27 and 37) requiring glutamine for growth produces derepressed levels of nitrogenase activity and synthesized catalytically inactive glutamine synthetase protein, as determined immunologically. A third class of glutamine-requiring, nitrogenase-derepressed mutants (strain SK-25 and 26) synthesizes neither a catalytically active glutamine synthetase enzyme nor an immunologically cross-reactive glutamine synthetase protein. 4. F-prime complementation analysis reveals that the mutant strains SK-25, 26, 27, 37 map in a segment of the Klebsiella chromosome corresponding to the region coding for glutamine synthetase. Since the mutant strains SK-27 and SK-37 produce inactive glutamine synthetase protein, it is concluded that these mutations map within the glutamine synthetase structural gene.     

Glutamate Dehydrogenase Reaction as a Source of Glutamic Acid in Synaptosomes

The role of the glutamate dehydrogenase reaction as a pathway of glutamate synthesis was studied by incubating synaptosomes with 5 mM 15NH4Cl and then utilizing gas chromatography-mass spectrometry to measure isotopic enrichment in glutamate and aspartate. The rate of formation of [15N]glutamate and [15N]aspartate from 5 mM 15NH4Cl was approximately 0.2 nmol/min/mg of protein, a value much less than flux through glutaminase (4.8 nmol/min/mg of protein) but greater than flux through glutamine synthetase (0.045 nmol/min/mg of protein). Addition of 1 mM 2-oxoglutarate to the medium did not affect the rate of [15N]glutamate formation. O2 consumption and lactate formation were increased in the presence of 5 mM NH3, whereas the intrasynaptosomal concentrations of glutamate and aspartate were unaffected. Treatment of synaptosomes with veratridine stimulated reductive amination of 2-oxoglutarate during the early time points. The production of ([15N]glutamate + [15N]aspartate) was enhanced about twofold in the presence of 5 mM beta-(+/-)-2-aminobicyclo [2.2.1]heptane-2-carboxylic acid, a known effector of glutamate dehydrogenase. Supplementation of the incubation medium with a mixture of unlabelled amino acids at concentrations similar to those present in the extracellular fluid of the brain had little effect on the intrasynaptosomal [glutamate] and [aspartate]. However, the enrichment in these amino acids was consistently greater in the presence of supplementary amino acids, which appeared to stimulate modestly the reductive amination of 2-oxoglutarate. It is concluded: (a) compared with the phosphate-dependent glutaminase reaction, reductive amination is a relatively minor pathway of synaptosomal glutamate synthesis in both the basal state and during depolarization; (b) NH3 toxicity, at least in synaptosomes, is not referable to energy failure caused by a depletion of 2-oxoglutarate in the glutamate dehydrogenase reaction; and (c) transamination is not a major mechanism of glutamate nitrogen production in nerve endings.     

Linking Diversity and Stable Isotope Fractionation in Ammonia-Oxidizing Bacteria

The link between similarity in amino acid sequence for ammonia monooxygenase (AMO) and isotopic discrimination for ammonia oxidation ( l AMO ) was investigated in g -subdivision ammonia-oxidizing bacteria. The isotope effects for ammonia oxidation in pure cultures of the nitrifying strains Nitrosomonas marina , Nitrosomonas C-113a, Nitrosospira tenuis , Nitrosomonas europaea , and Nitrosomonas eutropha ranged from 14.2 to 38.2. The differences in isotope effects could not be readily explained by differential rates of ammonia oxidation, transport of NH 4 + , or accumulation of NH 2 OH or N 2 O among the strains. The major similarities and differences observed in l AMO are, however, paralleled by similarities and differences in amino acid sequences for the f -subunit of AMO (AmoA). Robust differences in l AMO among nitrifying bacteria may be expected to influence the stable isotopic signatures of nitrous oxide (N 2 O) produced in various environments.     

Regulation of nitrogen fixation in Rhizobium spp. Isolation of mutants of Rhizobium trifolii which induce nitrogenase activity

This communication describes the isolation and characterization of mutants of Rhizobium trifolii which can induce nitrogenase activity in defined liquid medium. Two procedures were used for the isolation of these mutants from R. trifolii strain DT-6: (1) following chemical mutagenesis, slow growing mutants were selected which were unable to utilize NH+4 as sole source of nitrogen; (2) as spontaneous mutants resistant to the glutamate analogue L-methionine-DL-sulfoximine. Mutants (DT-71, DT-125) isolated by these procedures induced nitrogenase activity in the free-living state, whereas the parent strain lacked this property. Induction of nitrogenase activity in these mutants occurred during the late exponential phase of growth when the rate of protein synthesis was decreasing. The addition of NH+4 to a medium containing glutamate as the nitrogen-source resulted in a 50--70% reduction (repression?) of nitrogenase activity; in contrast, the rate of protein synthesis or the rate of respiration was not influenced by exogenous NH+4. Biochemical analysis showed that these mutants (strains DT-71 and DT-125) have defects in both nitrogen and carbon metabolism. The levels of glutamate synthase (both NADP+ -and NAD+ -dependent activities) and glutamate dehydrogenase (NAD+-dependent activity) were markedly lower. In addition, the mutants were found to have no detectable ribitol dehydrogenase or beta-galactosidase activity. These findings are discussed in relation to a mechanism of regulation of symbiotic nitrogen fixation.     

Bicarbonate Requirement for Elimination of the Lag Period of Hydrogenomonas eutropha

Carbon dioxide and oxygen concentrations have a profound effect on the lag period of chemoautotrophically grown Hydrogenomonas eutropha. Minimum lag periods and high growth rates were obtained in shaken flask cultures with a prepared gas mixture containing 70% H(2), 20% O(2), and 10% CO(2). However, excessively long lag periods resulted when the same gas mixture was sparged through the culture. The lag period was shortened in sparged cultures by decreasing both the pO(2) and the pCO(2), indicating that gas medium equilibration had not occurred in shaken cultures. The lag period was completely eliminated at certain concentrations of O(2) and CO(2). The optimum pO(2) was 0.05 atm, but the optimum pCO(2) varied according to the pH of the medium and physiological age of the inoculum. At pH 6.4, the pCO(2) required to obtain immediate growth of exponential, postexponential, and stationary phase inocula at equal specific rates was 0.02, 0.05, and 0.16 atm, respectively. With each 0.3-unit increase in the pH of the medium, a 50% decrease in the CO(2) concentration was needed to permit growth to occur at the same rate. The pCO(2) changes required to compensate for the pH changes of the medium had the net effect of maintaining a constant bicarbonate ion concentration. Initial growth of H. eutropha was therefore indirectly related to pCO(2) and directly dependent upon a constant bicarbonate ion concentration.     

Stimulation of Dopamine Release from Cultured Rat Mesencephalic Cells by Naturally Occurring Excitatory Amino Acids: Involvement of Both N-Methyl-D-Aspartate (NMDA) and Non-NMDA Receptor Subtypes

In rat mesencephalic cell cultures, L-glutamate at concentrations ranging from 100 microM to 1 mM stimulated release of [3H]dopamine that was attenuated by the non-N-methyl-D-aspartate (non-NMDA) receptor antagonist 6,7-dinitroquinoxalinedione, but not by the selective NMDA receptor antagonists (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine hydrogen maleate (MK-801; 10 microM) and 3-(2-carboxypiperazine-4-yl)propyl-1-phosphonate (300 microM). Even at 1 mM glutamate, this release was Ca2+ dependent. These observations suggest that the release was mediated by a non-NMDA receptor. Only release stimulated by a lower concentration (10 microM) of glutamate was inhibited by MK-801 (10 microM), indicating that glutamate at this concentration activates the NMDA receptor. By contrast, L-aspartate at concentrations of 10 microM to 1 mM evoked [3H]dopamine release that was completely inhibited by MK-801 (10 microM) and was also Ca2+ dependent (tested at 1 and 10 mM aspartate). Thus, effects of aspartate involved activation of the NMDA receptor. Sulfur-containing amino acids (L-homocysteate, L-homocysteine sulfinate, L-cysteate, L-cysteine sulfinate) also evoked [3H]dopamine release. Release evoked by submillimolar concentrations of these amino acids was attenuated by MK-801 (10 microM), indicating involvement of the NMDA receptor.(ABSTRACT TRUNCATED AT 250 WORDS)     

Control of synbiotic nitrogen fixation in Rhizobia. Regulation of NH4+ assimilation.

This communication is concerned with physiological, biochemical, and genetic studies of the regulation of ammonium (NH4+) assimilation by Rhizobia (root nodule bacteria) that infect leguminous plants. The major conclutions are (i) physiological studies show that Rhizobia are able to assimilate NH4+ for growth only when supplemented with certain organic nitrogen sources (e.g., L-aspartate, L-leucine, L-serine). Addition of as little as 2 mug/ml of L-aspartate supported growth on NH4+ as nitrogen source. In contrast, addition of glutamate in combination with NH4+-blocked NH4+ utilization; (ii) biochemical analysis show that glutamate synthase activity (NADP- and NAD-linked) is always present in cells capable of assimilating NH4+; also cells without glutamate synthase activity were found to be incapable of NH4+ utilization. Glutamate synthase levels were observed to fluctuate markedly depending on the available nitrogen source and on the growth stage of the culture; (iii) mutants were selected in which assimilation of NH4+ is no longer subject to inhibition (repression?) by glutamate. The levels of glutamate synthase activity (NADP-linked) (in the presence of glutamate) show approximately a two-fold increase over the level in the parent strain. The mutants no longer require supplementation with small amounts of organic nitrogen for growth in medium containing inorganic nitrogen (e.g., NH4+ or NO3-); (iv) these findings are discussed in relation to the working model of symbiotic nitrogen fixation recently proposed (O'Gara and Shanmugam (1976), Biochim. Biophys. Acta 437, 313--321).