Requirement of homocitrate for the transfer of a 49V-labeled precursor of the iron-vanadium cofactor from VnfX to nif-apodinitrogenase.

A vanadium- and iron-containing cluster has been shown previously to accumulate on VnfX in the Azotobacter vinelandii mutant strain CA11.1 (DeltanifHDKvnfDGK::spc). In the present study, we show the homocitrate-dependent transfer of (49)V label from VnfX to nif-apodinitrogenase in vitro. This transfer of radiolabel correlates with acquisition of acetylene reduction activity. Acetylene is reduced both to ethylene and ethane by the hybrid holodinitrogenase so formed, a feature characteristic of alternative nitrogenases. Structural analogues of homocitrate prevent the acetylene reduction ability of the resulting dinitrogenase. Addition of NifB cofactor (-co) or a source of vanadium (Na(3)VO(4) or VCl(3)) does not increase nitrogenase activity. Our results suggest that there is in vitro incorporation of homocitrate into the V-Fe-S cluster associated with VnfX and that the completed cluster can be inserted into nif-apodinitrogenase. The homocitrate incorporation reaction and the insertion of the cluster into nif-apodinitrogenase (alpha(2)beta(2)gamma(2)) do not require MgATP. Attempts to achieve FeV-co synthesis using extracts of other FeV-co-negative mutants were unsuccessful, showing that earlier steps in FeV-co synthesis, such as the steps requiring VnfNE or VnfH, do not occur in vitro.

synthesis of FeMo-co. The genes coding for the structural components of dinitrogenase (nifDK) are not required (5). FeMo-co is assembled elsewhere and then inserted into dinitrogenase. The metabolic product of the protein encoded by nifB is NifBco, a FeMo-co precursor that has been shown to be the iron and sulfur donor to the cofactor (6,7). NifB-co is most probably also a precursor of FeV-co and FeFe-co, because nifB is required for vanadium-dependent and iron only-dependent diazotrophic growth of A. vinelandii (8). The gene products of nifE and nifN form a heterotetrameric protein (NifNE) that has been shown to bind NifB-co (9) and is proposed to serve as a scaffold during cofactor synthesis (10). The counterparts of NifE and NifN in the vnf system (VnfE and VnfN) probably serve a similar function during the synthesis of FeV-co. The product of nifH (and vnfH), dinitrogenase reductase, has multiple functions in the nitrogenase system. Besides being the obligate electron donor to dinitrogenase during catalysis, it is also required for FeMo-co (FeV-co) biosynthesis (11)(12)(13). Its exact role in this process is not understood. The nifV gene codes for homocitrate synthase (14). Homocitrate is a structural component of FeMo-co and presumably FeV-co and FeFe-co, because nifV is required for full functionality of all three dinitrogenases (15). The products of nifX and its homologue vnfX have also been shown to be involved in FeMo-co and FeV-co synthesis, respectively, although their exact role remains to be established. A vanadium-and iron-containing cluster having similar EPR characteristics to FeV-co accumulates on VnfX during FeV-co synthesis (16). NifX stimulates the synthesis of FeMo-co in an in vitro system (17). Both proteins are able to bind NifB-co (16). 2 The complete biosynthetic pathways of FeMo-co and FeV-co have not yet been elucidated, and other not yet identified components have been shown to stimulate the synthesis of FeMo-co in vitro (17).
An in vitro FeMo-co synthesis system has been described (18). It typically involves mixing of extracts from two different mutants defective in the synthesis of the cofactor or a mutant extract complemented with the purified missing component(s). In vitro synthesis of FeV-co has not been accomplished yet. Here we report the homocitrate-dependent in vitro incorporation of a VnfX-associated V-Fe-S cluster into nifapodinitrogenase.

EXPERIMENTAL PROCEDURES
Materials-All materials used for growth media preparation were of analytical grade. 49 V (in 6 N HCl, 0.5-1.0 mCi/ml) was purchased from Los Alamos National Laboratories. Sodium orthovanadate, sodium metavanadate, homocitric acid, isocitric acid, and malic acid were from Sigma. Tris base and glycine were purchased from Fisher. Acrylamide/ bisacrylamide solution was obtained from Bio-Rad. Ammonium tetra-thiomolybdate was a gift from Dr. D. Coucouvanis.
Bacterial Strains and Culture Conditions-A. vinelandii strains CA12 (⌬nifHDK (19)), CA11.1 (⌬nifHDKvnfDGK::spc (20)), CA117.3 (⌬nifDK⌬nifB (21)), DJ42.48 (⌬nifENvnfE705::kan (22)), CA11.8 (⌬nifHDK⌬vnf 72 (23)), UW45 (nifB (8)), and CA11.6.82 (tungstentolerant, vnfD82::Tn5-B21⌬nifHDK Rif r (24)) used in this work have been described. The strains were grown in 20-liter carboys containing 15 liters of Burk's modified medium with 40 g/ml nitrogen in the form of ammonium acetate and 1 M NaVO 3 , or 1 mM Na 2 WO 4 , or 1 M Na 2 MoO 4 . The cultures were incubated at 30°C and aerated vigorously. They were monitored for the depletion of ammonium, following which derepression of nif or vnf genes was allowed for 4 -5 h. Cells were collected using a Pellicon cassette system equipped with a filtration membrane (0.45 m, Millipore Corporation) and frozen at Ϫ80°C. For growth in the presence of vanadium or tungsten, all glassware used to prepare the culture medium and for cell growth was washed with 4 N HCl and rinsed thoroughly with deionized water. Cultures in the presence of 49 V were grown in 1-liter flasks with 250 ml of growth medium containing NaVO 3 (0.1 M) spiked with radioactive 49 VCl 5 (0.05 Ci/ml) and 40 g/ml nitrogen in the form of ammonium acetate. Cells were grown overnight at 30°C with shaking, collected by centrifugation, and resuspended in nitrogen-free medium containing NaVO 3 and 49 VCl 5 (same concentrations as above). They were then incubated for 5 h at 30°C for derepression of the vnf-nitrogenase system. Cells were harvested by centrifugation and frozen at Ϫ80°C.
Buffer Preparation-All buffers were sparged with purified N 2 for 20 -30 min, and sodium dithionite was added to a final concentration of 1 mM. Buffers used for protein purification contained 0.5 g/ml leupeptin and 0.2 mM phenylmethylsulfonyl fluoride.
Cell-free Extracts and Sephadex G-25 Chromatography-Cell-free extracts were prepared by osmotic shock, as described earlier (25). For some experiments the extract of A. vinelandii strain CA11.1 grown in vanadium-containing medium was chromatographed on Sephadex G-25 to remove small molecular weight molecules. Forty ml of extract were loaded on a Sephadex G-25 column (2.5 ϫ 51 cm) equilibrated with 25 mM Tris-HCl (pH 7.4), 10% glycerol. All procedures were performed anaerobically.
In Vitro Cofactor Synthesis and Nitrogenase Assay-Stoppered 9-ml serum vials that had been washed with 4 N HCl and thoroughly rinsed with deionized water were used for the reactions. The assay consists of two phases. During the first phase (35 min) cofactor synthesis and insertion into apodinitrogenase are allowed to occur. The second phase is the reduction of acetylene by the nitrogenase holoenzyme formed during the first phase. The reactions were carried out under argon. The standard reaction mixture contained the following ingredients (first phase) in a total volume of 0.75 ml: 0.1 ml of 25 mM Tris-HCl (pH 7.4) containing 0.5 mM sodium dithionite, 20 l of 5 mM homocitrate (pH 8), 10 l of 1 mM Na 3 VO 4 , 0.2 ml of an ATP-regenerating mixture (27) containing 0.1 mM dithionite, NifB-co (0.75 nmol of iron), dinitrogenase reductase (52 g of protein), 0.2 ml each of the extracts to be tested (between 1.7 and 2.5 mg of protein), and, when stated, purified nifapodinitrogenase (34 g of protein). When reconstituted with an excess of purified FeMo-co, 34 g of purified nif-apodinitrogenase produced 36 nmol of ethylene/min, and the nif-apodinitrogenase present in 200 l of UW45 (W) extract produced 35 nmol of ethylene/min. To ensure that the quantity of nif-apodinitrogenase used in the assays was saturating, titration experiments were performed using a fixed amount of CA11.1 (V) extract (0.2 ml, 1.7 mg of protein) and increasing amounts of nifapodinitrogenase (purified or as extract of UW45 (W)) (data not shown). The amount of nif-apodinitrogenase (purified or as extract of UW45 (W)) used in the assays was saturating. For some control reactions 10 l of 1 mM Na 2 MoO 4 were added instead of Na 3 VO 4 . This mixture was incubated for 35 min at 30°C with shaking. After this incubation the second phase of the assay was initiated by injection of an additional 0.8 ml of ATP-regenerating solution (containing 4 mM dithionite) and 52 g of dinitrogenase reductase to each vial. The nitrogenase assay was started by injecting 0.5 ml of acetylene into the vial. The vials were incubated at 30°C for 30 min with shaking. The reaction was stopped by adding 0.1 ml of 4 N NaOH, and the ethylene and ethane formed were measured with a Shimadzu model GC8A gas chromatograph equipped with a Porapak N (Waters Associates) column. Activities are expressed as nanomoles of ethylene (or ethane) formed per min per assay. When studying the effect of ATP during the cofactor synthesis and insertion phase, 10 l of ammonium tetrathiomolybdate (1 mM in N-methylformamide, pH 8) (28) were added right after the first phase of the assay, and the vials were incubated for 10 min at room temperature before starting the second phase of the assay, to prevent further insertion of the cofactor into apodinitrogenase. Reactions to be analyzed by anoxic native gel electrophoresis only underwent the first phase of the reaction (cofactor synthesis and insertion), after which they were placed on ice. Aliquots (100 l) of the reaction mixture were subjected to anoxic native gel electrophoresis, as described below.
Anoxic Native Gel Electrophoresis and Phosphorimaging-The procedures for anoxic native gel electrophoresis and phosphorimaging have been described (6). Proteins were resolved on 7-16% acrylamide (37.5:1 acrylamide:bisacrylamide) and 0 -20% sucrose gradient gels. After electrophoresis for 15 h at 100 V, the gels were dried and exposed to a phosphor screen for 1-3 days. Screens were scanned using a Cyclone storage phosphor system (Packard Instruments).

RESULTS AND DISCUSSION
To investigate in vitro FeV-co synthesis, extracts of different mutant strains defective in FeV-co (or FeMo-co) synthesis were mixed and incubated in conditions similar to those that have been described for in vitro FeMo-co synthesis (18). When an extract of vanadium-grown A. vinelandii CA11.1, which lacks the structural components of vnf-nitrogenase, was complemented with an extract of tungsten-grown A. vinelandii UW45 (nifB Ϫ ) (in the presence of other components; see "Experimental Procedures"), acetylene reduction activity was observed (Table I, line 1). High activities were also observed when the extract of A. vinelandii UW45 was replaced by purified nifapodinitrogenase (Table I, line 2). Both the extract of UW45 (W) and the purified nif-apodinitrogenase were added in amounts that were saturating in the assay, as determined by titration (see "Experimental Procedures"). The enzyme responsible for the acetylene reduction activity observed is probably a hybrid of nif-dinitrogenase and FeV-co. The activity detected is characteristic of alternative nitrogenases, because ethane as well as ethylene were produced from acetylene. The ratio of ethylene to ethane produced (C 2 H 6 /C 2 H 4 ϫ 100) ranged from 2.2 to 2.9. These values are in the same range that has been described for vnf-nitrogenases (29,30) and for a hybrid of nif-dinitrogenase containing FeV-co (31). A control assay con-  (20 M), NifB-co (0.75 nmol of iron), NifH (52 g of protein), 0.2 ml of each extract (1.7-2.5 mg of protein), and nif-apodinitrogenase (34 g) when stated. Thirty-four g of purified nif-apodinitrogenase produced 36 nmol of ethylene/min when reconstituted with an excess of purified FeMo-co. The nif-apodinitrogenase present in 200 l of UW45 (W) extract produced 35 nmol of ethylene/min when reconstituted with an excess of purified FeMo-co. The amounts of nif-apodinitrogenase (purified or as extract of UW45 (W)) are saturating in the assays. Total volume equals 0.75 ml. (V) or (W) indicate that the cells used were grown in the presence of vanadium (V) or tungsten (W).

Mixture
Ethylene formed Ethane formed nmol/min nmol/min Transfer of 49 V from VnfX to nif-apodinitrogenase taining Mo and all factors necessary for in vitro FeMo-co synthesis showed high ethylene production activity but no detectable ethane, as expected (Table I, line 3).
Because the reaction mixture used contains all factors necessary for FeMo-co synthesis, except for Mo, it is important to show that the activity seen (or part of it) is not due to FeMo-co being synthesized using contaminating Mo present in the system. The level of activity arising from contaminating Mo present in the assay solutions and in the extract of UW45 grown in tungsten is very low, as can be seen in Table I, line 9. This control assay contains all components necessary for in vitro FeMo-co synthesis, except for Mo. A control experiment using an air-oxidized extract of vanadium-grown A. vinelandii CA11.1 and all other factors necessary for the reaction exhibited significantly reduced activity (Table I, Table I, lines 1 and 2 do not arise from contaminating Mo supplied by the extract of A. vinelandii CA11.1. Furthermore, when the assay was carried out with an extract of a tungstentolerant A. vinelandii strain that is incapable of Mo accumulation (CA11.6.82, tungsten-tolerant, vnfD82::Tn5-B21⌬nifHDK Rif r , (24)), similar levels of activity were observed (Table I, line 4).
To show that the cofactor present in this nitrogenase contains vanadium, the assay was carried out under standard conditions using purified nif-apodinitrogenase and an extract of A. vinelandii CA11.1 that had been grown in the presence of 49 V (a radioactive isotope of vanadium). The assay mixture was electrophoresed under anoxic native conditions, and the gel was analyzed by phosphorimaging. A transfer of label from VnfX to the position of dinitrogenase was observed (Fig. 1, lane  2). An extract of A. vinelandii CA12 (⌬nifHDK) grown in the presence of 49 V was loaded in Fig. 1, lane 4 as a standard to show the migration of vnf-dinitrogenase in the gel. 55 Fe-labeled nif-dinitrogenase obtained by in vitro FeMo-co synthesis using 55 Fe-NifB-co is shown in Fig. 1, lane 5. Both nitrogenases run as multiple bands in anoxic native gel conditions. The reason for this is unknown. It has been described previously that in A. vinelandii CA11.1 most of the 49 V accumulates on VnfX, as seen in Fig. 1, lane 1, in the form of a vanadium-containing Fe-S cluster (16).
Experiments in which a source of vnf-apodinitrogenase was provided instead of nif-apodinitrogenase were also conducted. Extracts of A. vinelandii CA117.3 (⌬nifDK⌬nifB) and CA11.8 (⌬nifHDKvnfH) were used as sources of vnf-apodinitrogenase. Unexpectedly, very low nitrogenase activity was observed in these experiments (Table I, lines 6 and 7). Accordingly, no 49 V was seen associated to dinitrogenase by phosphorimage analysis of anoxic native gels of these incubations (Fig. 1, lane 3). The reason for the failure to see nitrogenase activity in these conditions is currently unknown. All the factors necessary for the synthesis of FeV-co should be present in the extract of A. vinelandii CA11.1 grown in vanadium, because the addition of purified nif-apodinitrogenase alone is enough to see the activity. Thus, the lack of activity observed when using vnf-apodinitrogenase could be due to inefficient insertion of the cofactor. It is possible that an unknown factor required for insertion is labile and does not withstand the cell lysis or assay conditions. It has been shown that the product of vnfG is required for insertion of FeV-co into vnf-dinitrogenase (21). VnfG was present in the extracts used as a source of vnf-apodinitrogenase, as estimated by Western blot analysis (data not shown). What protein, if any, was facilitating the insertion of the newly formed FeV-co into nif-apodinitrogenase in the experiments described earlier is not known. VnfG was not present in the incubation mixtures containing nif-apodinitrogenase, because the mutant strain used (CA11.1) has a deletion in vnfDGK. The non-nif protein gamma has been shown to function as a chaperone that aids insertion of FeMo-co into nif-apodinitrogenase (26). Gamma is expressed under vnf conditions and was present in the extracts of A. vinelandii CA11.1, as well as part of the nif-apodinitrogenase used (␣ 2 ␤ 2 ␥ 2 ). A stable association of FeV-co with gamma has not been observed before and gamma is not able to replace VnfG effectively in the insertion of FeV-co into vnf-apodinitrogenase (21). It is possible that gamma can transiently associate with FeV-co and allow its insertion into nif-apodinitrogenase. Alternatively, some other protein or VnfX itself may be acting as the insertase in this system. Experiments with purified components where no gamma is present will provide insight into this problem.
Effect of Homocitrate Analogues-Because A. vinelandii CA11.1 is defective in the subunits for vnf-dinitrogenase and not in any of the genes known to encode proteins required for the biosynthesis of FeV-co, it seemed possible that the activity seen in Table I (lines 1, 2, and 4) was due to FeV-co present in this extract. It has been reported that the V-Fe-S cluster accumulating on VnfX in extracts of A. vinelandii CA11.1 contains no or very low levels of homocitrate (16). To test whether homocitrate incorporation into the unfinished cluster occurs in vitro, we studied the effect of homocitrate analogues on the reaction. These organic acids are able to replace homocitrate in the FeMo-co synthesis system (27,32), resulting in the synthesis of aberrant forms of FeMo-co that are known to exhibit altered substrate specificity (27,32). The use of isocitrate and malate in FeMo-co synthesis results in dinitrogenases with 8 -16% of the wild type acetylene reduction activities (27). When added in an excess in in vitro FeMo-co synthesis assays containing homocitrate, these organic acids cause a reduction in the level of the acetylene reduction activity of the resulting dinitrogenase. Isocitrate and malate were added to the assays containing an extract of vanadium-grown A. vinelandii CA11.1 and nif-apodinitrogenase (Table II). When no organic acid was added to the reaction, some residual activity was seen, presumably resulting from the homocitrate present in the extract. High activity was observed when 0.1 mM homocitrate (final concentration) was added. This concentration has been shown to be saturating for the in vitro FeMo-co synthesis reaction (32). In assay mixtures containing 0.1 mM homocitrate and isocitrate (5 and 10 mM), acetylene reduction activities were decreased by 86 and 89%, respectively. The highest concentration of malate used (20 mM) resulted in a 46% decrease in the acetylene reduction ability of nitrogenase. To ensure that the effect of the Transfer of 49 V from VnfX to nif-apodinitrogenase organic acids observed was not an inhibition of the coupled acetylene reduction assay, isocitrate and malate were added to control reactions immediately before the acetylene reduction assays were started (Table II, lines 10 -13). The highest concentrations used of malate and isocitrate only inhibited acetylene reduction by 10.4 and 5.9%, respectively. Thus, the low activities observed when adding the organic acids during the synthesis phase probably reflect the incorporation of the homocitrate analogues into FeV-co during the reaction.
Requirements for the Homocitrate Incorporation Reaction-To further characterize the reaction taking place when incubating the extract of vanadium-grown A. vinelandii CA11.1 with nif-apodinitrogenase, a crude extract was chromatographed on Sephadex G-25 to remove low molecular weight molecules. The desalted extract was used for the reac-tions described in Table III and Fig. 2. As expected, the complete system showed high acetylene reduction activity (Table  III, line 1) and a shift of the 49 V label to the position of nitrogenase in the anoxic native gel (Fig. 2, lane 2). No activity was observed when homocitrate was omitted from the reaction mixture (Table III, line 4). Accordingly, the appearance of the label on nitrogenase was dependent on the presence of homocitrate (Fig. 2, lane 3). 55 Fe-labeled nif dinitrogenase obtained by in vitro FeMo-co synthesis using 55 Fe-NifB-co is shown in Fig. 2, lane 6. The hexameric form of nif-apodinitrogenase (␣ 2 ␤ 2 ␥ 2 ), which is directly activable by FeMo-co, was used in these reactions. Apodinitrogenase can also exist in a tetrameric form (␣ 2 ␤ 2 ), which is not activable by FeMo-co. The maturation of the ␣ 2 ␤ 2 form into the ␣ 2 ␤ 2 ␥ 2 form requires ATP and dinitrogenase reductase (26, 33) but not the insertion of FeMo-co into the mature ␣ 2 ␤ 2 ␥ 2 form. In our system, ATP was found not to be necessary for the homocitrate incorporation reaction or for , and extract of A. vinelandii UW45 (nifB Ϫ) grown in tungsten medium.

TABLE II
Effect of homocitrate analogues on acetylene reduction activity by resulting nitrogenase In vitro FeV-co synthesis assays were carried out under standard conditions using purified nif-apodinitrogenase (34 g) and an extract of A. vinelandii CA11.1 grown in the presence of vanadium (1.7 mg of protein), as described in "Experimental Procedures." Thirty-four g of purified nif-apodinitrogenase produced 36 nmol of ethylene/min when reconstituted with an excess of purified FeMo-co. nif-apodinitrogenase is saturating in the assays. HC, homocitrate.  3 9.25 Ϯ 0.49 0.20 Ϯ 0.01 a The complete system contained the following: extract of vanadiumgrown A. vinelandii CA11.1 that had been desalted by Sephadex G-25 chromatography (1.6 mg of protein), nif-apodinitrogenase (34 g), homocitrate (HC) (0.1 mM), ATP-regenerating mix, Na 3 VO 4 (20 M), NifB-co (0.75 nmol of iron), NifH (52 g of protein). Thirty-four g of purified nif-apodinitrogenase produced 36 nmol of ethylene/min when reconstituted with an excess of purified FeMo-co. nif-apodinitrogenase is saturating in the assays. b n.d., not detected. c Ammonium tetrathiomolybdate (10 nmol per assay) was added to this reaction after the synthesis phase. The reaction was then incubated for 10 min at room temperature before adding MgATP, NifH, and acetylene for the substrate reduction phase of the assay.   (Table III, line 5; Fig. 2, lane 5). The omission of NifB-co from the incubation mixture did not have a significant effect on the observed activity (Table III, line 6), nor did the exclusion of Na 3 VO 4 or its replacement by VCl 3 (Table III, lines 7 and 8). Similarly, the presence of unlabeled Na 3 VO 4 in the reaction mixture did not affect the intensity of the 49 V label seen on nitrogenase (Fig. 2,  lane 4). These observations are consistent with the idea that the reaction occurring in vitro is the incorporation of homocitrate into the cluster associated with VnfX that already contains vanadium, iron, and sulfur. When the reaction was carried out using a desalted extract of A. vinelandii CA11.1 grown with a source of unlabeled vanadium, and 49 VCl 5 was included in the reaction mixture, no 49 V was observed associated with dinitrogenase in phosphorimages of anoxic native gels, despite high nitrogenase activities (data not shown). Thus, the in vitro FeV-co synthesis system seems to use only vanadium already associated with the cluster. It is also conceivable that incorporation of vanadium into the cluster occurs in vitro but that the system is only able to use the chemical form of vanadium present in the extract, which may be different from the form of vanadium supplied externally. However, desalted 49 V-labeled extracts show homocitrate-dependent incorporation of 49 V into nitrogenase (Fig. 2, lane 2), and thus, we conclude that the label is transferred from a macromolecule-bound form.
Complementation of Crude Extracts of Other FeV-co-negative Mutants-To assess whether earlier steps in the synthesis of FeV-co occur in vitro, we tried to complement extracts of mutant strains defective in gene products known to be involved in FeV-co synthesis with the missing components. VnfH (NifH) is known to be involved in FeV-co (FeMo-co) synthesis (12,13), although its exact role during the synthesis remains unknown. Thus, nifH vnfH double mutants are unable to synthesize FeMo-co and FeV-co. No acetylene reduction activities were seen when trying to complement an extract of a mutant strain defective in VnfH and NifH (A. vinelandii CA11.8) grown in vanadium medium with an extract of UW45 grown in tungsten medium, which contains NifH and nif-apodinitrogenase (Table  IV, line 1). The same results were obtained when purified apodinitrogenase and NifH were used instead of the extract of UW45 grown in tungsten (Table IV, line 3) or when a source of VnfH (an extract of CA117.3 grown in vanadium) was added to these assays (Table IV, lines 2 and 4). A mutant strain defective in VnfNE and NifNEX (A. vinelandii DJ42.48) was also used in complementation studies. Extracts of A. vinelandii DJ42.48 grown in vanadium medium were complemented either with purified NifNE and VnfX (in the presence of nif-apodinitrogenase) (Table IV, lines 7 and 8) or with an extract of A. vinelandii UW45 grown in tungsten medium and VnfX (Table IV, lines 5 and 6). The VnfX used in these assays was purified from a nifB Ϫ strain and does not have a cluster bound to it (16). The acetylene reduction activities observed in these conditions were not significantly above background. These results indicate that earlier steps in the synthesis of FeV-co do not occur in vitro under the assay conditions used in this study. They further strengthen the notion that the activities reported in Table I correspond to the addition of homocitrate to an already formed cluster that accumulates on VnfX in the mutant strain CA11.1. In contrast, in vitro synthesis of FeMo-co starting from NifB-co is routinely observed when using conditions similar to those tried here for FeV-co synthesis. The failure to obtain earlier steps in FeV-co synthesis in vitro is not currently understood.
Conclusions-The results of this study suggest that there is incorporation of homocitrate in vitro into the V-Fe-S cluster associated with VnfX in A. vinelandii CA11.1. The newly formed vanadium-containing cluster can be inserted into nifapodinitrogenase to yield an active enzyme that can convert acetylene to ethylene and ethane. Thus, incorporation of homocitrate into FeV-co would occur as a late step in the synthesis pathway of the cluster. More research needs to be done to identify what proteins are involved in the homocitrate incorporation reaction.