(Received for publication, February 24, 1997, and in revised form, June 12, 1997)
From the Department of Biochemistry and Center for the Study of Nitrogen Fixation, College of Agricultural and Life Sciences, University of Wisconsin-Madison, Madison, Wisconsin 53706
NIFH (the nifH gene product) has
several functions in the nitrogenase enzyme system. In addition to
reducing dinitrogenase during nitrogenase turnover, NIFH functions in
the biosynthesis of the iron-molybdenum cofactor (FeMo-co), and in the
processing of 2
2 apodinitrogenase 1 (a
catalytically inactive form of dinitrogenase 1 that lacks the FeMo-co)
to the FeMo-co-activatable
2
2
2 form. The
molybdenum-independent nitrogenase 2 (vnf-encoded) has a
distinct dinitrogenase reductase protein, VNFH. We investigated the
ability of VNFH to function in the in vitro biosynthesis of
FeMo-co and in the maturation of apodinitrogenase 1. VNFH can replace
NIFH in both the biosynthesis of FeMo-co and in the maturation of
apodinitrogenase 1. These results suggest that the dinitrogenase
reductase proteins do not specify the heterometal incorporated into the
cofactors of the respective nitrogenase enzymes. The specificity for
the incorporation of molybdenum into FeMo-co was also examined using the in vitro FeMo-co synthesis assay system.
The reduction of atmospheric N2 to ammonium by biological systems is catalyzed by the nitrogenase enzymes. The aerobe Azotobacter vinelandii harbors three genetically distinct nitrogenase enzymes that are regulated by the metal content of the growth medium, among other factors (1-3). Nitrogenases 1, 2, and 3 are encoded by the nif, vnf, and anf genes, respectively. nif-encoded nitrogenase 1 is a molybdenum-containing enzyme that is expressed in the presence of molybdenum. Expression of the vnf-encoded nitrogenase 2, a vanadium-containing enzyme, requires medium that is depleted in molybdenum and that contains vanadium. The anf-encoded nitrogenase 3 is expressed in medium deficient in both metals. All three nitrogenases are two-component metalloenzymes comprised of dinitrogenase and dinitrogenase reductase (1, 2). Dinitrogenase contains the active site metal center of the enzyme, and dinitrogenase reductase functions as the obligate electron donor to dinitrogenase during enzyme turnover in a MgATP- and reductant-dependent process (4, 5).
The active site of dinitrogenase 1, the iron-molybdenum cofactor (FeMo-co),1 is composed of molybdenum, iron, sulfur, and homocitrate ((R)-2-hydroxyl-1,2,4-butanetricarboxylic acid) in a 1:7:9:1 ratio (6-9). The biosynthesis of FeMo-co involves the products of at least six nif genes, including nifQ, nifB, nifV, nifE, nifN, and nifH (9-12). The nifQ gene product might be involved in the formation of a molybdenum-sulfur precursor to FeMo-co (13), and the nifV gene product encodes homocitrate synthase (9, 10).2 The product of NIFB, termed NifB-co, is an iron- and sulfur-containing precursor to FeMo-co (14, 15). Based on the amino acid sequence identity of NIFEN to NIFDK (the structural polypeptides of dinitrogenase 1), and the fact that NIFEN has been shown to bind NifB-co (15-17), NIFEN has been proposed to be a scaffold for FeMo-co assembly; however, the precise function of NIFEN in FeMo-co biosynthesis is unknown, as is the function of NIFH (dinitrogenase reductase 1).
In addition to being necessary for the biosynthesis of FeMo-co, the gene products of both nifV and nifB are required for the biosynthesis of the iron-vanadium cofactor (FeV-co) of dinitrogenase 2 and the putative iron-only cofactor (FeFe-co) of dinitrogenase 3 (18-21). Thus, homocitrate is presumed to be present as a component of FeV-co and FeFe-co, and NifB-co is believed to serve as the iron and sulfur donor to all three cofactors. Homologs of nifE and nifN have been identified in the vnf but not in the anf system (22); homologs of nifH exist in both molybdenum-independent systems (23, 24). Additional gene products required for FeV-co and FeFe-co biosynthesis have not been identified.
An in vitro system for the synthesis of FeMo-co that requires at least molybdate, homocitrate, an ATP-regenerating mixture, a source of reductant, NifB-co, NIFEN, and NIFH has been described (12, 14, 25, 26). The in vitro FeMo-co synthesis system utilizes molybdenum with high specificity as addition of 100-fold excess tungstate (a competitive inhibitor of the molybdenum transport system in Klebsiella pneumoniae) or vanadate do not significantly inhibit FeMo-co synthesis (12). The replacement of molybdenum with vanadium or iron in the cofactor during in vitro synthesis has not been achieved. The preferential incorporation of molybdenum into FeMo-co suggests that a component(s) involved in FeMo-co biosynthesis might exclusively select for molybdenum. The presence of a dinitrogenase reductase associated with each nitrogen fixation system makes that protein a likely candidate for specifying the heterometal incorporated into the respective cofactors of the nitrogenase enzymes.
NIFH has multiple roles in the nitrogenase 1 enzyme system. In addition
to MgATP-dependent electron transfer to dinitrogenase during substrate reduction, NIFH is required for the biosynthesis of
FeMo-co (10, 11) and for the maturation of apodinitrogenase 1 (a
catalytically inactive form of dinitrogenase 1 that lacks FeMo-co) to
its FeMo-co-activatable form (27, 28). In the latter process, NIFH is
required for the association of the protein (a
non-nif-encoded protein) (28) with
2
2 apodinitrogenase 1 to form the
FeMo-co-activatable
2
2
2
hexamer (27). Some altered forms of NIFH that are unable to function as
a reductant for nitrogenase-dependent substrate reduction
are fully functional in FeMo-co biosynthesis and in the maturation of
apodinitrogenase 1 (29-31), indicating that the characteristics of
NIFH that enable it to function in nitrogenase turnover are not
necessary for its role in the formation of active dinitrogenase.
In vivo studies by Joerger et al. (23) and Gollan
et al. (32) suggest that NIFH supports FeV-co synthesis and
that ANFH (the anfh gene product) supports FeMo-co
synthesis. We utilized the in vitro FeMo-co synthesis assay
system to definitively determine whether VNFH would function in FeMo-co
biosynthesis; the ability of VNFH to replace NIFH in the formation of
the FeMo-co-activatable 2
2
2 form of
apodinitrogenase 1 was also examined. Studies on the specificity of the
incorporation of molybdenum into FeMo-co are discussed.
DEAE-cellulose was a Whatman DE52 product. Sephacryl S-100 and the Mono Q anion exchange column were from Pharmacia Biotech Inc. The fast protein liquid chromatography instrument was from LKB. Sodium dithionite (DTH) was purchased from Fluka Chemicals. Sodium metavanadate (NaVO3, 99.995% purity), Tris base, and glycine were Fisher products. Acrylamide/bisacrylamide solution was obtained from Bio-Rad. All reagents used for A. vinelandii growth medium were of analytical grade or higher purity. Tetrathiomolybdate ((NH4)2MoS4) was a gift from D. Coucouvanis, and [K2(H2O)5][(VO2)2(R,S-homocitrate)2]·H2O was a gift from W. Armstrong (33). All other chemicals were from Sigma.
A. vinelandii Strains and Growth ConditionsA.
vinelandii strains DJ1030 (nifH
nifB)
(28), CA12 (
nifHDK) (34), UW45
(nifB[minus0]) (35), and CA117.30 (
nifDKB)
(36) have been described. All vessels used in preparing media and for
cell culture were rinsed thoroughly in 4 N HCl and then in
deionized water. Cultures (15 liters) of strain DJ1030 were grown in
20-liter polycarbonate carboys with vigorous aeration at 30 °C on
Burk's medium that lacked sodium molybdate and contained 10 µM NaVO3 (for derepression of the
vnf system) and 40 µg of nitrogen/ml as ammonium acetate. The cultures were monitored for depletion of ammonium, following which
they were derepressed for 4.5 h. The cells were concentrated using
a Pellicon cassette system equipped with a filtration membrane (0.45 µm, Millipore Corp., Bedford, MA) and were centrifuged. The cell
pellets were frozen in liquid N2 and stored at
80 °C. Strain DJ1030 was grown on Burk's medium containing 1 mM
sodium molybdate in place of NaVO3 for derepression of the
nif system. Strain UW45 was grown and derepressed on
tungsten-containing medium (molybdenum free) as described previously
(12). Strain CA117.30 was grown in 250-ml cultures on Burk's medium
containing 10 µM NaVO3; cells were
concentrated by centrifugation, and derepression was initiated (for
4 h) by suspending the cell pellets in nitrogen-free Burk's
medium. Cells were harvested by centrifugation and frozen as described
above. Cell-free extracts were prepared by the osmotic shock method
(6).
All buffers were sparged with purified N2 (and degassed on a gassing manifold where appropriate) for 10-30 min, and DTH was added to a final concentration of 1.7 mM. All buffers used in column chromatography contained 0.2 mM phenylmethylsulfonyl fluoride and 0.5 µg/ml leupeptin. Buffers used in fast protein liquid chromatography were filtered through a 0.45-µm filter. Tris-HCl was at pH 7.4 unless stated otherwise.
Purification of VNFHAll column chromatography steps except
for fast protein liquid chromatography were performed at 4 °C. VNFH
was purified from extract of strain DJ1030
(nifH
nifB, vnf-derepressed) with
modifications to the method described by Hales et al. (37).
One hundred fifty ml of cell-free extract (from 50 g of cell
paste) were applied to a 2.5 × 17-cm DEAE-cellulose column that
had been equilibrated in buffer containing 0.1 M NaCl in
0.025 M Tris-HCl, pH 7.4. Following application of the
extract, the column was washed with 2 bed volumes of buffer containing
0.125 M NaCl in 0.025 M Tris-HCl; VNFH was eluted using 0.22 M NaCl in 0.025 M Tris-HCl.
The DEAE-cellulose fraction was concentrated by ultrafiltration using a
XM100-A membrane, and the retentate (4 ml) was applied to a 2.5 × 78-cm Sephacryl S-100 column that had been equilibrated with 0.05 M NaCl in 0.025 M Tris-HCl. The column was
developed with the same buffer, and the VNFH-containing fractions that
exhibited the highest activity were concentrated by ultrafiltration
(described above) and purified further on a Mono Q anion exchange
column used in conjunction with a fast protein liquid chromatography
system. Two ml (6.8 mg of protein) of the VNFH-containing retentate
from the ultrafiltration cell was applied onto the Mono Q column that
had been equilibrated with 0.15 M NaCl in 0.025 M Tris-HCl. The column was washed with 1 bed volume of the
equilibration buffer, following which VNFH was eluted using a 20-ml
increasing linear gradient from 0.15 to 0.4 M NaCl (in
0.025 M Tris-HCl, pH 7.4). VNFH eluted with 0.32 M NaCl in 0.025 M Tris-HCl. Active fractions
were stored in 9-ml, serum-stoppered vials at
80 °C. The ability
of VNFH to transfer electrons to dinitrogenase 1 was tested using the acetylene reduction assay for nitrogenase activity, and the results were consistent with the published results. VNFH was equally effective as NIFH in transferring electrons to dinitrogenase 1, consistent with
the results of Chisnell et al. (34).
FeMo-co was prepared in N-methylformamide as
described previously (6). The reactions were performed in 9-ml,
serum-stoppered vials that were repeatedly evacuated, flushed with
argon, and rinsed with 0.3 ml of 0.025 M Tris-HCl
containing 1.7 mM DTH. The following components were added
to the vials in the order indicated: 100 µl of 0.025 M
Tris-HCl; 200 µl of an ATP-regenerating mixture (containing 3.6 mM ATP, 6.3 mM MgCl2, 51 mM phosphocreatine, 20 units/ml creatine phosphokinase, and
6.3 mM DTH); 200 µl (3.8 mg protein) of extract of strain
DJ1030 (nifH
nifB,
nif-derepressed) as a source of
2
2 apodinitrogenase 1 and the
protein; and 10-50 µl (0.1 mg of protein) of the appropriate
dinitrogenase reductase. The vials were incubated for 10 min at room
temperature to allow the formation of
2
2
2 apodinitrogenase 1. One hundred µl of anoxic 50% glycerol were added to the reactions to
be analyzed by native PAGE, and these vials were placed on ice. Ten
µl of a solution containing an excess of FeMoco were added to the
remaining vials, which were incubated for 10 min at room temperature
during which
2
2
2
apodinitrogenase 1 was activated by FeMo-co to form dinitrogenase 1. Fifty nmol of (NH4)2MoS4 (prepared
in N-methylformamide containing 1.7 mM DTH) were
added to the vials to prevent further FeMo-co insertion into
apodinitrogenase 1. Activity of the newly reconstituted dinitrogenase 1 was monitored by the C2H2 reduction assay for
nitrogenase (12). (NH4)2MoS4 was
excluded in certain control reactions, and 0.1 mg of the appropriate
dinitrogenase reductase (that used in the insertion phase of the assay)
was added in place of 0.1 mg of NIFH normally added during the
C2H2 reduction phase of the assay (12).
Nine-ml serum vials were
repeatedly evacuated, flushed with argon, and rinsed with buffer
containing 1.7 mM DTH. Components were added to the vials
in the following order: 100 µl of 0.025 M Tris-HCl, 10 µl of 1 mM Na2MoO4, 20 µl of 5 mM homocitrate (that had been treated with base to cleave
the lactone, pH 8.0), and 200 µl of the ATP-regenerating mixture
(defined above). The vials were incubated at room temperature for
10-20 min. Two hundred µl of extract (~3.8 mg protein) of either
DJ1030 (nifH
nifB, nif-derepressed) or CA12 (
nifHDK,
nif-derepressed), 25 µl of a solution containing NifB-co,
and 10-50 µl (0.1 mg protein) of the appropriate dinitrogenase
reductase were added to the vials. The vials were incubated at 30 °C
for 30-90 min. Following this incubation, 100 µl of anoxic 50%
glycerol were added to the reactions to be analyzed by anoxic native
PAGE, and these vials were placed on ice. Five nmol of
(NH4)2MoS4 (prepared as described
above) were added to the remaining vials to prevent further FeMo-co
synthesis during the subsequent C2H2 reduction
phase of the assay. The activity of the newly formed dinitrogenase 1 was monitored by the C2H2 reduction assay.
(NH4)2MoS4 was excluded from
certain reactions to which 0.1 mg of the appropriate dinitrogenase
reductase (that used in the synthesis phase of the assay) was added in
place of 0.1 mg of NIFH normally added during the
C2H2 reduction phase of the assay.
Proteins were resolved on anoxic native gels with a 7-14% acrylamide (37.1% acrylamide, 1% bisacrylamide) and 0-20% sucrose gradient. The electrophoresis buffer was N2-sparged, 65 mM Tris-glycine (pH 8.5) containing 1.7 mM DTH. Gels were pre-electrophoresed for at least 60 min at 120 V for initial reduction, and proteins were electrophoresed for 1920 V-h (at 120 V) at 4 °C. One hundred µl of the reaction mixtures were applied onto the gel.
Antibodies and Immunoblot AnalysisPolyclonal antibodies to
NIFH and the protein were raised in rabbits (the anti-
protein
antibodies were prepared and made available by Drs. Mary Homer and Gary
Roberts). Immunoblotting and developing procedures have been described
(38). The native gels were equilibrated in transfer buffer for at least
15 min prior to blotting.
Protein concentrations of cell-free extracts and purified proteins were measured using the bicinchoninic acid method (39).
The
association of the protein with
2
2
apodinitrogenase 1 to form the
2
2
2 FeMo-co-activable
species requires the presence of NIFH and nucleotide (27, 28), as
diagrammed in Reaction 1.
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Apodinitrogenase 1 |
Apodinitrogenase 1 |
Dinitrogenase 1 |
(catalytically active) |
![]() |
|
To confirm the results of the FeMo-co insertion assays, we employed
anoxic, native PAGE to monitor the association of the protein with
2
2 apodinitrogenase 1 in extracts of
strain DJ1030 (
nifH
nifB,
nif-derepressed) in the presence of nucleotide and the
different dinitrogenase reductase proteins. Fig.
1, an immunoblot of an anoxic, native gel
(developed with antibody to the
protein), illustrates that VNFH
functions in the association of the
protein with
2
2 apodinitrogenase 1 (Fig. 1, lane
3). These results are consistent with the activities observed in
the FeMo-co insertion assays testing the different dinitrogenase
reductase proteins (Table II).
|
The high degree of amino acid sequence identity between NIFH and VNFH (91%) (1) is consistent with the effectiveness of VNFH in both substrate reduction (when complemented with dinitrogenase 1) and in the maturation of apodinitrogenase 1. The domain(s) of NIFH required for both the above functions are quite likely highly conserved in VNFH. At present, the role(s) of the dinitrogenase reductase protein in the maturation of apodinitrogenase 1 remains under investigation.
Ability of VNFH to Function in in Vitro FeMo-co SynthesisVNFH was tested in the in vitro FeMo-co synthesis assay in place of NIFH (Table I). VNFH typically exhibited 25-30% of the FeMo-co synthesis activity (in our fixed time assay) observed with an equivalent level of NIFH, despite exhibiting similar levels of activity in the C2H2 reduction assay. Addition of increasing levels of VNFH and increasing the time allowed for in vitro FeMo-co synthesis did not result in a linear increase in activity (data not shown). The limiting step(s) in the assay is not the maturation of apodinitrogenase 1, because VNFH functions as effectively as NIFH in the maturation process (discussed above). The reasons for the lower level of FeMo-co synthesis observed with VNFH are not known. It is possible that VNFH is unable or slow to dissociate from a nif protein(s) with which it interacts during the course of FeMo-co synthesis, thus limiting further turnover of the protein(s) involved.
Homer et al. (28) demonstrated that the protein dimer
(present in extracts of A. vinelandii strains unable to
synthesize FeMo-co) monomerized upon associating with FeMo-co, and thus
it was possible to employ the monomerization of the
protein
(detected by anoxic native PAGE) as an alternate assay for the
completion of FeMo-co synthesis. Thus, FeMo-co synthesized in
vitro in reaction mixtures containing an extract of strain CA12
(
nifHDK, nif-derepressed) would accumulate on
the
protein (resulting in the monomerization of the
protein
dimer) due to the absence of apodinitrogenase 1 in extracts of this
strain. Fig. 2 is an immunoblot
(developed with antibody to the
protein) of an anoxic native gel
that demonstrates the results of this study. When dinitrogenase
reductase is excluded from the in vitro FeMo-co synthesis
reaction, the
dimer and a slow migrating species of
that is
uncharacterized (indicated by X on Fig. 2) are observed
(Fig. 2, lane 1); the dimeric form of the
protein is
observed in extracts of strains that are impaired in FeMo-co
biosynthesis (33). That both NIFH and VNFH support FeMo-co biosynthesis
is illustrated by the monomerization of the
protein observed as the
faster migrating
protein-FeMo-co form in reactions that included
NIFH or VNFH (Fig. 2, lanes 2 and 3).
Does dinitrogenase reductase specify the heterometal contained in the nitrogenase cofactors? Two lines of evidence suggest that the dinitrogenase reductases do not specify or select against the heterometal that is incorporated into the cofactors of the nitrogenase enzymes: 1) the ability of VNFH to function in in vitro FeMo-co synthesis (albeit less effectively than NIFH), and 2) the observation by Joerger et al. (23) that NIFH supported vanadium-dependent diazotrophic growth of an A. vinelandii strain containing a deletion in the vnfH gene, indicating that, in vivo, NIFH functions in FeV-co biosynthesis. Gollan et al. (32) demonstrated the in vivo synthesis and incorporation of FeMo-co into the dinitrogenase 3 polypeptides of a Rhodobacter capsulatus strain containing deletions in the nifHDK genes; the synthesis of FeMo-co in the absence of a nifH gene suggests that ANFH most likely replaced NIFH in the synthesis of FeMo-co. Our results demonstrating the ability of VNFH to function in the in vitro biosynthesis of FeMo-co suggest that the dinitrogenase reductase protein quite likely does not select against the incorporation of molybdenum into FeV-co and FeFe-co.
The Specificity for Molybdenum of the in Vitro FeMo-co Synthesis SystemCofactor structures of the three nitrogenases are proposed to be essentially similar with vanadium and iron atoms replacing the molybdenum atom in FeV-co and FeFe-co, respectively (2, 21, 40). The requirement of the nifB and nifV gene products for the biosynthesis of all three cofactors suggests that certain steps in the biosynthesis of FeMo-co are shared in the biosynthetic pathways of all three cofactors. Although FeV-co is largely uncharacterized, extended x-ray absorption fine structure studies on dinitrogenase 2 indicate that FeV-co is similar in structure to FeMo-co with the octahedral vanadium atom surrounded by 3 oxygen atoms and 3 sulfur atoms as is the molybdenum atom in FeMo-co (41). Other similarities between FeMo-co and FeV-co include the ability to extract FeV-co into N-methylformamide (20) and its probable ligation to the dinitrogenase 2 polypeptide via the conserved cysteine and histidine residues (analogous to Cys-275 and His-442 of NIFD) that ligate FeMo-co to dinitrogenase 1 (8, 23).
To determine whether the FeMo-co synthesis system would utilize
vanadium and iron in the synthesis of FeV-co and FeFe-co, respectively,
we tested various vanadium- and iron-containing compounds in place of
molybdenum in the in vitro FeMo-co synthesis assay. Extract
of A. vinelandii strain UW45 (nifB,
tungsten-grown) was used as a source of all the nif-encoded proteins necessary for the synthesis of FeMo-co. Active dinitrogenase 1 was formed only when molybdenum (in the form of
Na2MoO4) was included in the in
vitro reactions (Table II). Molybdenum added to in
vitro FeMo-co synthesis reactions in the form of
(NH4)2MoO2S2, K2MoO3S, and MoS2 also supported
in vitro FeMo-co synthesis (data not shown). Vanadium added
in the form of NaVO3, V2O5,
VCl3,VOPO4, or
[K2(H2O)5][(VO2)2(R,S-homocitrate)2]·H2O
did not produce active dinitrogenase 1. Similar results were obtained
when iron (in the form of FeCl3 and Fe(II)NO3)
was included in the assay. Several possibilities might account for
these results. The FeMo-co synthesis machinery might indeed
discriminate against vanadium and iron; however, in vivo
studies demonstrating the ability of NIFEN and NIFH to support
vanadium-dependent diazotrophy suggest that certain nif proteins required for FeMo-co biosynthesis do function
in FeV-co biosynthesis in vivo (22, 23). Vanadium and iron
might not be in their correct oxidation states or precursor forms
necessary for incorporation into the cofactor under the in
vitro assay conditions.
We employed cell-free extracts of strain CA117.30
(nifDKB) that was derepressed on vanadium to determine
whether FeV-co could be synthesized under conditions similar to those
used to synthesize FeMo-co in vitro. When extract of
CA117.30 (
nifDKB, vnf-derepressed) was used as
a source of vnf-encoded proteins in in vitro
reactions containing vanadium (in the form of NaVO3,
V2O5, VCl3, VOPO4, or
K2(H2O)5][(VO2)2(R,S-homocitrate)2]·H2O),
homocitrate, ATP (in the form of an ATP-regenerating mixture), and
NifB-co, formation of active dinitrogenase 2 was not observed (Table
II). Varying the nucleotides included in the reactions, the pH of the
reaction mixture, and addition of partially purified apodinitrogenase 1 and NIFEN (in case of limiting levels of apodinitrogenase 2 and VNFEN
in the vnf-derepressed extracts) to certain reactions also produced negative results. Clearly, the in vitro conditions
under which FeMo-co is synthesized are inadequate for the synthesis of
FeV-co. As discussed above, the conversion of vanadium to the form
required for its incorporation into FeV-co might not occur in
vitro; alternatively, intermediates in the FeV-co biosynthetic pathway might be unstable under our cell-breakage and assay conditions. These observations suggest that steps and precursors unique to the
synthesis of FeV-co quite likely exist. The identification of
additional vnf genes and the characterization of phenotypes of strains carrying lesions in vnf genes might enable the
elucidation of steps involved in the biosynthesis of FeV-co.
We thank Gary Roberts and Priya Rangaraj for
helpful discussions and for critically reading this manuscript. We
thank Mary Homer and Gary Roberts for making anti- protein
antibodies available to us. Sandra Grunwald is gratefully acknowledged
for providing purified R. rubrum NIFH.