The vbs genes that direct synthesis of the siderophore vicibactin in Rhizobium leguminosarum: their expression in other genera requires ECF sigma factor RpoI.

A cluster of eight genes, vbsGSO, vbsADL, vbsC and vbsP, are involved in the synthesis of vicibactin, a cyclic, trihydroxamate siderophore made by the symbiotic bacterium Rhizobium leguminosarum. None of these vbs genes was required for symbiotic N2 fixation on peas or Vicia. Transcription of vbsC, vbsGSO and vbsADL (but not vbsP) was enhanced by growth in low levels of Fe. Transcription of vbsGSO and vbsADL, but not vbsP or vbsC, required the closely linked gene rpoI, which encodes an ECF sigma factor of RNA polymerase. Transfer of the cloned vbs genes, plus rpoI, to Rhodobacter, Paracoccus and Sinorhizobium conferred the ability to make vicibactin on these other genera. We present a biochemical genetic model of vicibactin synthesis, which accommodates the phenotypes of different vbs mutants and the homologies of the vbs gene products. In this model, VbsS, which is similar to many non-ribosomal peptide synthetase multienzymes, has a central role. It is proposed that VbsS activates L-N5-hydroxyornithine via covalent attachment as an acyl thioester to a peptidyl carrier protein domain. Subsequent VbsA-catalysed acylation of the hydroxyornithine, followed by VbsL-mediated epimerization and acetylation catalysed by VbsC, yields the vicibactin subunit, which is then trimerized and cyclized by the thioesterase domain of VbsS to give the completed siderophore.

Alleviation of aluminum toxicity to Rhizobium leguminosarum bv. viciae by the hydroxamate siderophore vicibactin.

Acid rain solubilises aluminum which can exert toxic effects on soil bacteria. The root nodule bacterium Rhizobium leguminosarum biovar viciae synthesises the hydroxamate siderophore vicibactin in response to iron limitation. We report the effect of vicibactin on the toxicity of aluminum(III) to R. leguminosarum and kinetic studies on the reaction of vicibactin with Al(III) and Fe(III). Aluminum (added as the nitrate) completely inhibited bacterial growth at 25 microM final concentration, whereas the preformed Al-vicibactin complex had no effect. When aluminum and vicibactin solutions were added separately to growing cultures, growth was partly inhibited at 25 microM final concentration of each, but fully inhibited at 50 microM final concentration of each. Growth was not inhibited at 50 microM Al and 100 microM vicibactin, probably reflecting the slow reaction between Al and vicibactin; this results in some aluminum remaining uncomplexed long enough to exert toxic effects on growth, partly at 25 microM Al and vicibactin and fully at 50 microM Al and vicibactin. At 100 microM vicibactin and 50 microM Al, Al was complexed more effectively and there was no toxic effect. It was anticipated that vicibactin might enhance the toxicity of Al by transporting it into the cell, but the Al-vicibactin complex was not toxic. Several explanations are possible: the Al-vicibactin complex is not taken up by the cell; the complex is taken up but Al is not released from vicibactin; Al is released in the cell but is precipitated immediately. However, vicibactin reduces the toxicity of Al by complexing it outside the cell.

Differential effects on N(2) binding and reduction, HD formation, and azide reduction with alpha-195(His)- and alpha-191(Gln)-substituted MoFe proteins of Azotobacter vinelandii nitrogenase.

In contrast to the wild-type MoFe protein, neither the alpha-195(Asn) nor the alpha-191(Lys) MoFe protein catalyzed N(2) reduction to NH(3), when complemented with wild-type Fe protein. However, N(2) was bound by the alpha-195(Asn) MoFe protein and inhibited the reduction of both protons and C(2)H(2). The alpha-191(Lys) MoFe protein did not interact with N(2). With the alpha-195(Asn) MoFe protein, the N(2)-induced inhibition of substrate reduction was reversed by removing the N(2). Surprisingly, even though added H(2) also relieved N(2) inhibition of substrate reduction, the alpha-195(Asn) MoFe protein did not catalyze HD formation under a N(2)/D(2) atmosphere. This observation is the first indication that these two reactions have different chemical origins, prompting a revision of the current hypothesis that these two reactions are consequences of the same nitrogenase chemistry. A rationale that accounts for the dichotomy of the two reactions is presented. The two altered MoFe proteins also responded quite differently to azide. It was a poor substrate for both but, in addition, azide was an electron-flux inhibitor with the 195(Asn) MoFe protein. The observed reactivity changes are correlated with likely structural changes caused by the amino acid substitutions and provide important details about the interaction(s) of N(2,) H(2), D(2), and azide with Mo-nitrogenase.

Azotobacter vinelandii nitrogenases with substitutions in the FeMo-cofactor environment of the MoFe protein: effects of acetylene or ethylene on interactions with H+, HCN, and CN-.

Wild-type and three altered Azotobacter vinelandii nitrogenase MoFe proteins, with substitutions either at alpha-195(His) (replaced by alpha-195(Asn) or alpha-195(Gln)) or at alpha-191(Gln) (replaced by alpha-191(Lys)), were used to probe the interactions of HCN and CN(-), both of which are present in NaCN solutions at pH 7.4, with nitrogenase. The first goal was to determine how added C(2)H(2) enhances the rate of CH(4) production from HCN reduction by wild-type nitrogenase. In the absence of C(2)H(2), wild-type Mo-nitrogenase showed a declining total electron flux, which is an overall measure of all products formed, as the NaCN concentration was increased from 1 to 5 mM, whereas the rates of both CH(4) and NH(3) production increased with increasing NaCN concentration. The NH(3) production rate exceeded the CH(4) production rate up to 5 mM NaCN, at which point they became equal. The "excess NH(3)" likely arises from the two-electron reduction of HCN to CH(2)=NH, some of which is released and hydrolyzed to HCHO plus NH(3). With added C(2)H(2), the rate of CH(4) production increased but only until it equaled that of NH(3) production, which remained unchanged. In addition, total electron flux was decreased even more at each NaCN concentration by C(2)H(2). The increased CH(4) production did not arise from the added C(2)H(2). The lowered total electron flux with C(2)H(2) present would decrease the affinity of the enzyme for HCN, making it a poorer competitor for the binding site. Thus, less CH(2)=NH would be displaced, more CH(2)=NH would undergo the full six-electron reduction, and the rate of CH(4) production would be enhanced. A second goal was to gain mechanistic insight into the roles of the amino acid residues in the alpha-subunit of the MoFe protein at positions alpha-191 and alpha-195 in substrate reduction. At 5 mM NaCN and in the presence of excess wild-type Fe protein, the specific activity for CH(4) production by the alpha-195(Asn), alpha-195(Gln), and alpha-191(Lys) MoFe proteins was 59%, 159%, and 6%, respectively, of that of wild type. For the alpha-195(Asn) MoFe protein, total electron flux decreased with increasing NaCN concentration like wild type. However, the rates of both CH(4) and NH(3) production were maximal at 1 mM NaCN, and they remained unequal even at 5 mM NaCN. With the alpha-195(Gln) MoFe protein, the rates of production of both CH(4) and NH(3) were equal at all NaCN concentrations, and total electron flux was hardly affected by changing the NaCN concentration. With the alpha-191(Lys) MoFe protein, the rates of both CH(4) and NH(3) production were very low, but the rate of NH(3) production was higher, and both rates slowly increased with increasing NaCN concentration. A hypothesis, which is based on the varying apparent affinities of the altered MoFe proteins for HCN and CN(-), is advanced to explain the higher rate of NH(3) production versus the rate of CH(4) production and the effect of increasing NaCN concentration on electron flux to products. A new method for CH(3)NH(2) quantification showed that all four MoFe proteins produced CH(3)NH(2). Added CO significantly inhibited both CH(4) and NH(3) production from HCN with all MoFe proteins except for the alpha-191(Lys) MoFe protein, which still manifested its very low rate of NH(3) production but without CH(4) production. All of the MoFe proteins responded differently to the addition of C(2)H(2) to reactions containing NaCN. With the alpha-195(Asn) MoFe protein, added C(2)H(2) decreased the rates of both CH(4) and NH(3) production, but the rate of NH(3) production decreased much less. C(2)H(2) also exacerbated the inhibition of electron flux. With the alpha-195(Gln) MoFe protein, added C(2)H(2) decreased the rates of both CH(4) and NH(3) production substantially and about equally. C(2)H(2) also eliminated the slight decrease in total electron flux that was caused by NaCN. Added C(2)H(2) hardly affected the alpha-191(Lys) MoFe protein. (ABSTRACT TRUNCA

Azotobacter vinelandii nitrogenases containing altered MoFe proteins with substitutions in the FeMo-cofactor environment: effects on the catalyzed reduction of acetylene and ethylene.

Altered MoFe proteins of Azotobacter vinelandii Mo-nitrogenase, with amino acid substitutions in the FeMo-cofactor environment, were used to probe interactions among C(2)H(2), C(2)H(4), CO, and H(2). The altered MoFe proteins used were the alpha-195(Asn) or alpha-195(Gln) MoFe proteins, which have either asparagine or glutamine substituting for alpha-histidine-195, and the alpha-191(Lys) MoFe protein, which has lysine substituting for alpha-glutamine-191. On the basis of K(m) determinations, C(2)H(2) was a particularly poor substrate for the nitrogenase containing the alpha-191(Lys) MoFe protein. Using C(2)D(2), a correlation was shown between the stereospecificity of proton addition to give the products, cis- and trans-C(2)D(2)H(2), and the propensity of nitrogenase to produce ethane. The most extensive loss of stereospecificity occurred with nitrogenases containing either the alpha-195(Asn) or the alpha-191(Lys) MoFe proteins, which also exhibited the highest rate of ethane production from C(2)H(2). These data are consistent with the presence of a common ethylenic intermediate on the enzyme, which is responsible for both ethane production and loss of proton-addition stereochemistry. C(2)H(4) was not a substrate of the nitrogenase with the alpha-191(Lys) MoFe protein and was a poor substrate of the nitrogenases incorporating either the wild-type or the alpha-195(Gln) MoFe protein, both of which had a low V(max) and high K(m) (120 kPa). Ethylene was a somewhat better substrate for the nitrogenase with the alpha-195(Asn) MoFe protein, which exhibited a K(m) of 48 kPa and a specific activity for C(2)H(6) formation from C(2)H(4) 10-fold higher than the others. Neither the wild-type nitrogenase nor the nitrogenase containing the alpha-195(Asn) MoFe protein produced cis-C(2)D(2)H(2) when turned over under trans-C(2)D(2)H(2). These results suggest that the C(2)H(4)-reduction site is affected by substitution at residue alpha-195, although whether the effect is related to the substrate-reduction site directly or is mediated through disturbance of the delivery of electrons/protons is unclear. Ethylene inhibited total electron flux, without uncoupling MgATP hydrolysis from electron transfer, to a similar extent for all four A. vinelandii nitrogenases. This observation indicates that this C(2)H(4) flux-inhibition site is remote from the C(2)H(4)-reduction site. Added CO eliminated C(2)H(4) reduction but did not fully relieve its electron-flux inhibition with all four A. vinelandii nitrogenases, supporting the suggestion that electron-flux inhibition by C(2)H(4) is not directly connected to C(2)H(4) reduction. Thus, C(2)H(4) has two binding sites, and the presence of CO affects only the site at which it binds as a substrate. When C(2)H(2) was added, it also eliminated C(2)H(6) production from C(2)H(4) and also did not relieve electron-flux inhibition fully. Thus, C(2)H(2) and C(2)H(4) are likely reduced at the same site on the MoFe protein. Two schemes are presented to integrate the results of the interactions of C(2)H(2) and C(2)H(4) with the MoFe proteins.

Legume root nodule bacteria and acid pH.

In 1984 the Australian Wool Research Trust Fund called for expressions of interest in projects directed at using the developing techniques of molecular biology for application to agricultural problems. With our interests in legume root nodule bacteria and their physiology, we felt that the problems for legume nodulation and N2 fixation posed by soils which were already acid, or which were rapidly acidifying, required just such attention. Further, the finding body's request coincided with the highly successful introduction into Western Australian agriculture of acid-tolerant strains of the medic-nodulating bacteria Sinorhizobium meliloti originating from acid soils on Sardinia (see below). The existence of such strains made it obvious that acid tolerance was a genetically determined trait, and provided invaluable biologically diverse material with which to work. The biological bases for that trait of acid tolerance were totally obscure, and many remain so, but the following account provides some light in the darkness. The research that we have done since in pursuit of explanations for acid tolerance have been funded first by the Wool Research Trust Fund and the Rural Credits Development Fund, and later by the Australian Research Council, and we here record our appreciation for their support.

Reversible alkaline inactivation of lignin peroxidase involves the release of both the distal and proximal site calcium ions and bishistidine co-ordination of the haem.

Phanerochaete chrysosporium lignin peroxidase isoenzyme H2 (LiP H2) exhibits a transition to a stable, inactive form at pH 9.0 with concomitant spectroscopic changes. The Söret peak intensity decreases some 55% with a red shift from 408 to 412 nm; the bands at 502 nm and 638 nm disappear and the peak at 536 nm increases. The EPR spectrum changes from a signal typical of high spin ferric haem to an exclusively low spin spectrum with g=2.92, 2.27, 1.50. These data indicate that the active pentaco-ordinated haem is converted into a hexaco-ordinated species at alkaline pH. Room temperature near-IR MCD data coupled with the EPR spectrum allow us to assign the haem co-ordination of alkali-inactivated enzyme as bishistidine. Re-acidification of the alkali-inactivated enzyme to pH 6 induces further spectroscopic changes and generates an irreversibly inactivated species. By contrast, a pH shift from 9.0 to 6.0 with simultaneous addition of 50 mM CaCl(2) results in the recovery of the initial activity together with the spectroscopic characteristics of the native ferric enzyme. Incubating with 50 mM CaCl(2) at a pH between 6.0 and 9.0 can also re-activate the enzyme. Divalent metals other than Ca(2+) do not result in restoration of activity. Experiments with (45)Ca indicate that two tightly bound calcium ions per enzyme monomer are lost during inactivation and reincorporated during subsequent re-activation, consistent with the presence of two structural Ca(2+) ions in LiP H2. It is concluded that both the structural Ca(2+) ions play key roles in the reversible alkaline inactivation of LiP H2.

Problems of adverse pH and bacterial strategies to combat it.

This chapter aims to survey the problems faced by bacteria found in environments of adverse pH, to review strategies used to combat those problems and to ask how those strategies are implemented. At acid or alkaline pH, bacteria are challenged not just by excess of H+ or OH- but also by excess of metal ions (aluminium, heavy metals at acidic pH, Na+ at alkaline pH), as well as shortages. Bacteria attempt to maintain their intracellular pH by minimizing membrane permeability to H+ and other ions, buffering the cytoplasm, ameliorating the external pH through catabolism or selective substrate utilization, and developing ionic pumping systems. The amelioration of pH depends on the availability of substrate, and is unlikely in most naturally stressful environments. Ion pumping is expensive energetically, although the cost to growth is unknown. The response to adverse pH involves sensing systems and responsive regulatory systems. The adaptive acid tolerance response is now well known in and other bacteria, but is there a widespread adaptive alkali tolerance response? What and where are the sensors? Whether they sense intracellular pH, extracellular pH or delta pH is unclear, although an external sensory input seems essential. Is there one major sensory system responsive to pH or multiple systems with back-up mechanisms? What and where are the regulators? Is there one central regulator controlling all the responses or are there cascades of responses?

Acid tolerance in root nodule bacteria.

Biological nitrogen fixation, especially via the legume Rhizobium symbiosis, is important for world agriculture. The productivity of legume crops and pastures is significantly affected by soil acidity; in some cases it is the prokaryotic partner that is pH sensitive. Growth of Rhizobium is adversely affected by low pH, especially in the 'acid stress zone'. Rhizobia exhibit an adaptive acid tolerance response (ATR) that is influenced by calcium concentration. Using Tn5-mutagenesis, gusA fusions and 'proteome' analysis, we have identified a range of genes that are essential for growth at low pH (such as actA, actP, exoR, actR and actS). At least three regulatory systems exist. The two-component sensor-regulator system, actSR, is essential for induction of the adaptive ATR. Two other regulatory circuits exist that are independent of ActR. One system involves the low pH-induced regulator gene, phrR, which may control other low pH-regulated genes. The other circuit, involving a regulator that is yet unidentified, controls the expression of a pH-regulated structural gene (lpiA). We have used pH-responsive gusA fusions to identify acid-inducible genes (such as lpiA), and then attempted to identify the regulators of these genes. The emerging picture is of a relatively complex set of systems that respond to external pH.

Effects on substrate reduction of substitution of histidine-195 by glutamine in the alpha-subunit of the MoFe protein of Azotobacter vinelandii nitrogenase.

Studies of the substrate-reducing capabilities of an altered nitrogenase MoFe protein (alpha-195(Gln) instead of alpha-195(His)) from a mutant of Azotobacter vinelandii show, contrary to an earlier report [Kim, C.-H., Newton, W. E., and Dean, D. R. (1995) Biochemistry 34, 2798-2808], that the alpha-195(Gln) MoFe protein can reduce N2 to NH3 but at a rate that is <2% of that of the wild type. The extent of effective binding of N2 by this altered MoFe protein, as monitored by the inhibition of H2 evolution, is markedly increased as temperature is lowered but virtually eliminated at 45 degreesC. This inhibition of H2 evolution results in an increase in the ATP:2e- ratio, i.e., the number of molecules of MgATP hydrolyzed for each electron pair transferred to substrate, from ca. 5 (the wild-type level) at 45 degreesC to nearly 25 at 13 degreesC. Like wild-type nitrogenase, the N2 inhibition of H2 evolution reaches a maximum at an Fe protein:MoFe protein molar ratio of ca. 2.5, suggesting that a highly reduced enzyme may not be necessary for N2 binding. N2 binding to the alpha-195(Gln) MoFe protein retains a hallmark of the wild type by producing HD under a mixed N2/D2 atmosphere. The rate of HD production and the fraction of total electron flow allocated to HD are similar to those for wild-type nitrogenase under the same conditions. However, the electrons forming HD do not come from those normally producing NH3 (as occurs in the wild type) but are equivalent to those whose evolution as H2 had been inhibited by N2. N2 also inhibits C2H2 reduction catalyzed by the alpha-195(Gln) nitrogenase. This inhibition is relieved by added H2, resulting in a lowering of the elevated ATP:2e- ratio to that found under Ar. With solutions of NaCN, which contain both the substrate, HCN, and the inhibitor, CN-, reduction of HCN is not impaired with the alpha-195(Gln) nitrogenase, but the inhibition by CN- of total electron flow to substrate, which is observed with the wild-type MoFe protein, is completely absent. Unlike that of the catalyzed reduction of H+, HCN, or C2H2, the extent of azide reduction to either N2 or N2H4 is markedly decreased (to 5-7% of that of the wild type) with the alpha-195(Gln) nitrogenase. Azide, like N2, inhibits H2 evolution and increases the ATP:2e- ratio. Both effects are freely reversible and abolished by CO. Added D2 does not relieve either effect, implying that N2 produced from N3- is not the inhibitory species. The correlation between the extremely low rates of reduction for both N2 and azide by the alpha-195(Gln) nitrogenase and their common ability to inhibit H2 evolution suggests that alpha-histidine-195 may be an important proton conductor to the FeMo cofactor center and specifically required for reduction of these two substrates.

A helicase gene (helO) in Rhizobium meliloti WSM419.

A 2.8 kb BamHI DNA fragment adjacent to a BamHI fragment containing actR-actS (a sensor/regulator pair required for low pH tolerance in Rhizobium meliloti WSM419) was cloned and sequenced. A computer predicted protein of 821 amino acids, designated HelO, showed extensive similarity with 'DEAH' motif helicases. Expression of helO was higher at pH 7.0 than pH 5.8 and it did not require the product of the actR gene. Inactivation of helO by insertion of a omega interposon at codon 40 did not affect nodulation, growth or tolerance to low pH, high temperature, osmolarity or elevated levels of copper or zinc.

Acid tolerance in Rhizobium meliloti strain WSM419 involves a two-component sensor-regulator system.

An acid-sensitive mutant, TG5-46, derived from Rhizobium meliloti WSM419 by Tn5 mutagenesis, fails to grow below pH 6.0 whereas the parent strain grows at pH 5.7. The DNA sequence of a 2.2 kb rhizobial DNA region flanking Tn5 in TG5-46 contains two open reading frames, ORF1 (designated actS) and ORF2 (designated actR), having high similarity to the sensor-regulator pairs of the two-component systems involved in signal transduction in prokaryotes. Insertion of an omega interposon into actS in R. meliloti WSM419 resulted in an acid-sensitive phenotype. A DNA fragment from the wild-type complemented the acid-sensitive phenotype of RT295 (ActS-) and TG5-46 (ActR-), while fragments containing only actR or actS complemented TG5-46 and RT295, respectively. The presence of multiple copies of actR complemented not only TG5-46 but also RT295. Cloning DNA upstream from actR and actS into a broad-host-range lacZ expression vector and measuring beta-galactosidase activities showed that both genes are constitutively expressed regardless of the external pH. Genomic DNA from all strains of R. meliloti, but no other bacteria tested, hybridized with an actRS probe at high stringency. These data implicate a two-component sensor-regulator protein pair in acid tolerance in R. meliloti and suggest their involvement in pH sensing and/or response by these bacteria.

Cloning and sequencing show that 4-hydroxybenzoate hydroxylase (PobA) is required for uptake of 4-hydroxybenzoate in Rhizobium leguminosarum.

Mutants of Rhizobium leguminosarum bv. viciae MNF300 and R. leguminosarum bv. trifolii WU95 unable to accumulate 4-hydroxybenzoate lack 4-hydroxybenzoate hydroxylase. The capacity of these mutants to take up and grow on 4-hydroxybenzoate was restored by a 2.0 kb EcoRI-PstI DNA fragment. This contained only one ORF which had over 60% DNA sequence similarity with the structural gene for 4-hydroxybenzoate hydroxylase (pobA) from Pseudomonas spp. and Acinetobacter. Reported effects of metabolic inhibitors and substrate analogues on the apparent uptake of 4-hydroxybenzoate have now been shown to be due to their direct effect on 4-hydroxybenzoate hydroxylase. We propose that uptake of 4-hydroxybenzoate is via a metabolic 'drag' mechanism dependent on the activity of the pobA gene product.

The molybdenum and vanadium nitrogenases of Azotobacter chroococcum: effect of elevated temperature on N2 reduction.

During the reduction of N2 by V-nitrogenase at 30 degrees C, some hydrazine (N2H4) is formed as a product in addition to NH3 [Dilworth and Eady (1991) Biochem. J. 277, 465-468]. We show here the following. (1) That over the temperature range 30-45 degrees C the apparent Km for the reduction of N2 to yield these products is the same, but increases from 30 to 58 kPa of N2. On increasing the temperature from 45 degrees C to 50 degrees C, little change occurred in the rate of reduction of protons to H2; the rate of N2H4 production increased, but the rate of NH3 formation decreased 7-fold. (2) Temperature-shift experiments from 42 to 50 degrees C or from 50 to 42 degrees C showed that this selective loss of the ability to reduce N2 to NH3 was reversible. The effects we observe are consistent with the existence of different conformers of the VFe-protein at the two temperatures, that predominating at 50 degrees C being largely unable to reduce N2 to ammonia. (3) Measurement of the ratio between H2 evolution and N2 reduced to NH3 at N2 pressures up to 339 kPa for both Mo- and V-nitrogenases gave limiting H2/N2 values of 1.13 +/- 0.13 for Mo-nitrogenase and 3.50 +/- 0.03 for V-nitrogenase. Since for Mo-nitrogenase our measured value for the ratio at 339 kPa is the same as that derived by Simpson and Burris [(1984) Science 224, 1095-1097] at 5650 kPa, there appears to be little or no divergence from the predictions based on the apparent Km for N2. These data then suggest that there may be a fundamentally different mechanism for N2 binding to V-nitrogenase compared with Mo-nitrogenase. (4) We did not detect any N2H4 as a product of N2 reduction by Mo-nitrogenase over the temperature range investigated; however, at 50 degrees C this system reduced acetylene (C2H2) to yield some ethane (C2H6), in addition to ethylene (C2H4), a reaction normally associated with Mo-independent nitrogenases.

Impaired nodule function in Medicago polymorpha L. infected with alfalfa mosaic virus.

The effects of alfalfa mosaic virus (AMV) on growth, nodule formation and nodule function in the annual burr medic, Medicago polymorpha cv. Circle Valley, were investigated in glasshouse pot experiments. Systemically-infected plants from AMV-infected seed produced 21% less shoot dry weight and accumulated 24% less fixed nitrogen in shoots than healthy plants when harvested 53 d after germination. At day 75, infected plants showed similar shoot dry weight losses (22%), but the quantity of nitrogen fixed fell by only 15%. At day 53, soluble sugar, starch and bacteroid concentrations in nodules were unaffected by AMV infection, but nitrogenase specific activity was decreased by 25% and soluble amino acids by 20%. Although AMV infection resulted in no differences in the number of nodules formed in the first 11 d after germination or at any harvest thereafter, nodule mass was decreased by 23% for virus-infected plants at day 53. However this difference disappeared by day 75. Growth of AMV-infected plants was decreased probably because of impaired N2 fixation with nodule function affected rather than nodulation. Increased nodule mass relative to plant weight in virus-infected plants, seen at day 75, implied some degree of compensation for the limitation in N2 -fixing capacity. ELISAs for AMV antigen indicated that nodules were active sites of virus multiplication.

Correction for creatine interference with the direct indophenol measurement of NH3 in steady-state nitrogenase assays.

Creatine was identified as a major source of interference with the direct phenol/hypochlorite colorimetric determination of ammonia in nitrogenase reaction mixtures. A method is described for removing other compounds which inhibit color development and for compensating for the interference produced by creatine. This method avoids time-consuming microdiffusion and also routinely makes available the efficiency of ATP hydrolysis coupled to substrate reduction (ATP/2e ratio) with N2 as a reducible substrate. Using this method we determined values for this ratio at 30 degrees C of 4.87 +/- 0.03 during the reduction of protons to H2 and 7.16 +/- 0.14 during the reduction of N2 by the vanadium-containing nitrogenase of Azotobacter chroococcum.

Siderophore and organic acid production in root nodule bacteria.

Nineteen strains of root nodule bacteria were grown under various iron regimes (0.1, 1.0 and 20 microM added iron) and tested for catechol and hydroxamate siderophore production and the excretion of malate and citrate. The growth response of the strains to iron differed markedly. For 12 strains (Bradyrhizobium strains NC92B and 32H1, B. japonicum USDA110 and CB1809, B. lupini WU8, cowpea Rhizobium NGR234, Rhizobium meliloti strains U45 and CC169, Rhizobium leguminosarum bv viciae WU235 and Rhizobium leguminosarum bv trifolii strains TA1, T1 and WU95) the mean generation time showed no variation with the 200-fold increase in iron concentration. In contrast, in Bradyrhizobium strains NC921, CB756 and TAL1000, B. japonicum strain 61A76 and R. leguminosarum bv viciae MNF300 there was a 2-5 fold decrease in growth rate at low iron. R. meliloti strains WSM419 and WSM540 showed decreased growth at high iron. All strains of root nodule bacteria tested gave a positive CAS (chrome azurol S) assay for siderophore production. No catechol-type siderophores were found in any strain, and only R. leguminosarum bv trifolii T1 and bv viciae WU235 produced hydroxamate under low iron (0.1 and 1.0 microM added iron). Malate was excreted by all strains grown under all iron regimes. Citrate was excreted by B. japonicum USDA110 and B. lupini WU8 in all iron concentrations, while Bradyrhizobium TAL1000, R. leguminosarum bv viciae MNF300 and B. japonicum 61A76 only produced citrate under low iron (0.1 and/or 1.0 microM added iron) during the stationary phase of growth.

Microscopic evidence on how iron deficiency limits nodule initiation in Lupinus angustifolius L.

Lupins (Lupinus angustifolius L. ev. Yandee), grown in solution culture, have been used to study the sites and process of infection by Bradyrhizobium sp. (Lupinus) and the impairment of nodulation by iron deficiency. Infection leading to nodulation occurred in an area of epidermal cells either lacking root hairs or with very young root hairs at the time of inoculation. Cells aged 13 h or over appeared not to be infected. Infection was initiated in the outer cortex. Rare, short infection threads were evident on day 4 after inoculation, 2 d after the initial division of cortical cells resulting from the bradyrhizobial inoculation. Bacteria had been released into the cytoplasm of cortical cells within 5 d after inoculation. Bacteroids multiplied in the cytoplasm, segregated passively and spread in the infection zones by repeated division of the invaded cells. Under iron deficiency, initial cell division occurred in the outer cortex of host roots, as in iron-sufficient plants after inoculation. Iron deficiency then limited further division of cortical cells. Only a few surviving infection sites developed nodules with normal structure but development was much slower than in iron-sufficient plants.

Hydrazine is a product of dinitrogen reduction by the vanadium-nitrogenase from Azotobacter chroococcum.

During the enzymic reduction of N2 to NH3 by Mo-nitrogenase, free hydrazine (N2H4) is not detectable, but an enzyme-bound intermediate can be made to yield N2H4 by quenching the enzyme during turnover [Thorneley, Eady & Lowe (1978) Nature (London) 272, 557-558]. In contrast, we show here that the V-nitrogenase of Azotobacter chroococcum produces a small but significant amount of free N2H4 (up to 0.5% of the electron flux resulting in N2 reduction) as a product of the reduction of N2. The amount of N2H4 formed increased 15-fold on increasing the assay temperature from 20 degrees C to 40 degrees C. Activity cross-reactions between nitrogenase components of Mo- and V-nitrogenases showed that the formation of free N2H4 was associated with the VFe protein. These data provide the first direct evidence for an enzyme intermediate at the four-electron-reduced level during the reduction of N2 by V-nitrogenase.