Reduction of short chain alkynes by a nitrogenase α-70 Ala-substituted MoFe protein

Article abstract of DOI:10.1039/b107336b

The Mo-dependent nitrogenase comprises two component proteins, called the Fe protein and MoFe protein, which together catalyze the nucleotide-dependent reduction of N₂ to NH₃. Nitrogenase also catalyzes the reduction of a variety of other multiply bonded substrates including the short chain alkyne acetylene Substrate reduction is known to occur at an Fe ₇S₉Mo: homocitrate cluster, designated FeMo-cofactor. Despite 40 years of intensive research and knowledge of the macromolecular structures of the nitrogenase component proteins from several different sources, the mechanism for the binding, activation and reduction of N₂ is still unclear. Based on amino acid substitution studies for those residues that provide the first shell of non-covalent interactions with FeMo-cofactor we previously targeted a specific 4Fe-4S face of FeMo-cofactor approached by the MoFe protein a-subunit residues α-70^Val and α-96 ^Arg as providing the substrate binding and reduction site. In the present work, support for this hypothesis was obtained by showing that substitution of the α-70^Val residue by α-70^Ala relaxes constraints within the substrate-binding pocket so that effective reduction of the short chain alkynes, propargyl alcohol and propyne, which are not effectively reduced by the wild type enzyme, is now permitted. The Royal Society of Chemistry 2002.

Full text of DOI:10.1039/b107336b

View Article Online / Journal Homepage / Table of Contents for this issue
Reduction of short chain alkynes by a nitrogenase
Ala
ꢀ
-70 -substituted MoFe protein†
Suzanne M. Mayer,* Walter G. Niehaus and Dennis R. Dean
Department of Biochemistry-Fralin Biotechnology Center, Virginia Tech, Blacksburg, Virginia,
2
4061, USA. E-mail: sumayer@vt.edu
Received 14th August 2001, Accepted 25th September 2001
First published as an Advance Article on the web 18th January 2002
The Mo-dependent nitrogenase comprises two component proteins, called the Fe protein and MoFe protein, which
together catalyze the nucleotide-dependent reduction of N to NH . Nitrogenase also catalyzes the reduction of a
2
3
variety of other multiply bonded substrates including the short chain alkyne acetylene. Substrate reduction is known
to occur at an Fe S Mo : homocitrate cluster, designated FeMo-cofactor. Despite 40 years of intensive research and
7
9
knowledge of the macromolecular structures of the nitrogenase component proteins from several different sources,
the mechanism for the binding, activation and reduction of N is still unclear. Based on amino acid substitution
2
studies for those residues that provide the first shell of non-covalent interactions with FeMo-cofactor we previously
Val
targeted a specific 4Fe-4S face of FeMo-cofactor approached by the MoFe protein α-subunit residues α-70 and
Arg
α-96 as providing the substrate binding and reduction site. In the present work, support for this hypothesis was
Val
Ala
obtained by showing that substitution of the α-70 residue by α-70 relaxes constraints within the substrate-
binding pocket so that effective reduction of the short chain alkynes, propargyl alcohol and propyne, which are
not effectively reduced by the wild type enzyme, is now permitted.
8
Introduction
certain cases by EPR spectroscopy, see Fisher et al. for a recent
example, only occurs at redox levels of the enzyme below the
resting state, and this requires the addition of Fe protein and
MgATP under turnover conditions. Although spectroscopic
analysis of nitrogenase under catalytic conditions has
Nitrogenase is a two-component metalloenzyme that catalyzes
the MgATP-dependent reduction of N . It is also able to reduce
2
a variety of other multiply bonded substrates, the most familiar
one being acetylene, which can be reduced by two electrons to
proven useful in the formulation of a kinetic description of
1
,2
yield ethylene. In the absence of other substrates, protons are
9–12
nitrogenase,
this approach has provided no mechanistic
reduced by nitrogenase to yield H . During catalysis the 60 kDa
Fe protein dimer (component II) delivers electrons to the 250
2
details about where or how substrates interact with FeMo-
cofactor. Finally, because substrates only interact transiently
with FeMo-cofactor, and at a redox level below the resting state,
crystallographic analysis of the enzyme complex in a substrate-
bound form has proven so far to be intractable. Because of
these inherent difficulties, formulation of plausible modes
for substrate interaction with the metal–sulfur surface of
FeMo-cofactor has been confined to theoretical calculations
and geometrical considerations. A brief but by no means
complete summary of a few of these models is described below.
Deng and Hoffmann used extended Hückel calculations
to identify a variety of possible N binding sites on FeMo-
cofactor. Three of the models showed N replacing one of the
bridging S ligands within FeMo-cofactor, three involved N2
binding to a 4Fe-4S face, one modeled N in the center of the
trigonal six-iron prism and the last modeled the replacement of
a terminal Mo-ligand with end-on N binding. These investiga-
tors favored a model where N interacts with two adjacent iron
atoms of the cluster. Dance used density function calculations
kDa α β MoFe protein (component I) which provides the
2
2
substrate reduction site. Initial intercomponent electron trans-
fer is thought to proceed from the solvent exposed [4Fe-4S]
cluster of the Fe protein to an [8Fe-7S] cluster located within
the MoFe protein, called the P-cluster, one of which is located
between each αβ-subunit pair. Electrons are subsequently
delivered from the P-cluster to the substrate-binding site
provided by
a
[7Fe-9S-Mo-Homocitrate] cluster called
3
FeMo-cofactor (see Christiansen et al. for a recent review).
One FeMo-cofactor is contained within each MoFe protein
α-subunit and each MoFe protein αβ unit is considered to
comprise an individual catalytic entity. FeMo-cofactor is
constructed from [4Fe-3S] and [3Fe-3S-Mo] subfragments
joined by three bridging sulfides with homocitrate attached
to the Mo atom through its 2-hydroxy and 2-carboxyl groups,
2
2
2
4–6
2
Fig. 1.
2
How do substrates interact with and become activated at
the metal–sulfur surface of FeMo-cofactor? This question has
proven difficult to answer for several reasons. First, although
intact FeMo-cofactor can be separated from purified MoFe
13
based on a symmetric [8Fe-9S] cluster model that is geometric-
ally analogous to the FeMo-cofactor metal–sulfur core. He
proposed that the under-coordinated Fe atoms at a 4Fe-4S face
of FeMo-cofactor are likely to play an important role in sub-
strate binding whereas adjacent sulfur atoms are involved in
7
protein, FeMo-cofactor removed from its polypeptide matrix
does not effectively reduce any substrate. Second, no significant
substrate interaction with FeMo-cofactor has been detected
for the MoFe protein in its as-isolated resting state. Rather,
productive substrate interaction, which can be detected in
proton transfer. The model he favored involves N binding at
2
14,15
a single Fe atom.
Sellmann et al. proposed a model where a
structural rearrangement of FeMo-cofactor occurs so that
a particular face of FeMo-cofactor accommodates substrate
binding and activation. This model suggests that a two-electron
reduction of FeMo-cofactor occurs prior to N2 binding at
which time a Fe–S–Fe bridge from the intact cluster dissociates
and instead ligates with nearby N or O donors from the
†
Based on the presentation given at Dalton Discussion No. 4, 10–13th
January 2002, Kloster Banz, Germany.
Research supported by National Institutes of Health Grant R01-
GM59087.
8
02
J. Chem. Soc., Dalton Trans., 2002, 802–807
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b107336b

View Article Online
20
impaired in its ability to reduce acetylene. Based on the
location and nature of the substituting residue that gives rise to
Gly
this phenotype (MoFe protein α-69 residue substituted by
Ser
α-69 ), we proposed that the 4Fe-4S face of FeMo-cofactor
Val
Arg
capped by α-70 and α-96 within the MoFe protein provides
20,21
the substrate binding site (Fig. 1).
Namely, we suggested
Ser
that the α-69 substitution has an indirect effect on acetylene
binding within the altered protein by impacting the dynamic
Val
movement of the side-chains of either or both α-70 and
Arg
α-96 during catalysis. As a way to test this hypothesis and
to determine if it is possible to relax the substrate selectivity
Val
of nitrogenase, we substituted the α-70 residue within the
Ala
MoFe protein by α-70 and determined the physiological and
catalytic consequences of this substitution.
Results
Val
Ala
Substitution of the MoFe protein ꢀ-70 residue by ꢀ-70 results
in physiogical sensitivity to propargyl alcohol
Val
A mutant A. vinelandii strain having the α-70 residue substi-
Ala
tuted by α-70 was constructed using site-directed mutagenesis
and gene replacement techniques. The strain producing the α-
Ala
7
0 -substituted MoFe protein (designated DJ1310) was
capable of normal diazotrophic growth and was therefore not
compromised in its ability to reduce N . To determine if con-
2
Ala
straints on substrate interaction were relaxed for the α-70
MoFe protein, the effect of adding propargyl alcohol
(
propargyl-OH) to the growth medium under nitrogen-fixing
conditions was evaluated for DJ1310 and the parental wild type
strain (Fig. 2). Propargyl-OH was chosen for this analysis
Fig. 1 (A) The FeMo-cofactor metal center excluding the protein
ligands and homocitrate. The specific 4Fe-4S face we are targeting is
labeled according to the deposited Av1 coordinates in the Protein Data
Bank (3MIN) and colored as follows: Fe atoms = bright green, S atoms
Fig. 2 Effect of propargyl-OH on diazotrophic growth of wild type A.
vinelandii and strain DJ1310. Strain DJ1310 produces an altered MoFe
protein having the α-70 residue substituted by α-70 . Cells are
spread on minimal medium petri plates having no added source of fixed
nitrogen. The petri plate on the left has strain DJ1310 (A) and wild type
Val
Ala
=
bright yellow. The rest of the metal center is colored as follows: Fe
atoms = dark green, S atoms = dark yellow, Mo = magenta. (B)
Structural organization of the FeMo-cofactor and its polypeptide
environment in relationship to the P-cluster. The helices connecting the
(
B) spread on minimal medium containing no fixed nitrogen source and
Val
Arg
no added propargyl-OH. The petri plate on the right has strain DJ1310
(C) and wild type (D) spread on minimal medium containing no fixed
nitrogen source with 6 mM propargyl-OH added. Photographs of the
petri plates were taken after one week of growth. Notice there is
no diazotrophic growth of strain DJ1310 in the presence of 6 mM
propargyl-OH (C).
P-cluster to α-70 and α-96
are shown as blue ribbons and the
Val
Arg
His
α-70 , α-96 and α-195 residues capping the specific 4Fe-4S face
that is being targeted are labeled. Atom colors are as follows: Mo =
magenta, Fe = green, S = yellow, C = grey, O = red and N = blue
4
3
spheres). Figures produced by Swiss-PDBViewer v3.7b2.
polypeptide environment. N subsequently binds in a bridging
2
because it is a relatively small substituted-acetylene molecule
that is also a water-soluble liquid at ambient temperature. The
addition of 6 mM propargyl-OH to the growth medium
completely inhibited diazotrophic growth of strain DJ1310 but
had little or no effect on diazotrophic growth of the wild type
strain. The addition of 6 mM propargyl-OH had no effect on
the growth of either strain when a source of fixed nitrogen was
also added to the growth medium, indicating that the effect is
mode between the two ‘open’ Fe atoms. In this model the role
of the Mo atom in catalysis is suggested to involve fine-tuning
16,17
the redox potential of the cluster.
Szilagyi et al. targeted
the Mo-site and its ligand environment using a hybrid density
function method (B3LYP). In line with a model previously
18
proposed by Grönberg et al. they suggested rearrangement of
the homocitrate from bidentate to monodentate coordination
19
as a way to accommodate N binding at the Mo atom.
2
specific for physiological N fixation.
2
As an alternative but complementary strategy to these
theoretical approaches we have taken a direct experimental
approach for identification of the substrate interaction site by
manipulating the polypeptide environment of FeMo-cofactor
using amino acid substitution. Towards this end we recently
described a genetic selection which resulted in the production
of an altered nitrogenase MoFe protein from Azotobacter
Propargyl-OH inhibition of proton reduction, N reduction and
2
Ala
acetylene reduction catalyzed by the ꢀ-70 MoFe protein
To determine the basis for propargyl-OH inhibition of diazo-
trophic growth by DJ1310, the substituted MoFe protein from
this strain was purified and the effect of propargyl-OH on the
reduction of other substrates examined. It is known that for
vinelandii that is able to effectively reduce N but is significantly
2
J. Chem. Soc., Dalton Trans., 2002, 802–807
803

View Article Online
nitrogenase catalysis a mixture of different substrates will com-
pete with each other for the available reducing equivalents
without affecting the overall rate of electron flux. Thus, the
Ala
interaction of propargyl-OH with the α-70 MoFe protein was
first evaluated by its ability to inhibit either proton reduction or
N reduction. These analyses (Fig. 3) show that propargyl-OH is
2
Fig. 4 Lineweaver–Burk plot showing competitive inhibition by
Ala
propargyl-OH on acetylene reduction catalyzed by the α-70 MoFe
protein. No addition (solid circles), plus 1.5 mM propargyl-OH
(
squares), plus 3.0 mM propargyl-OH (triangles), plus 5.0 mM
propargyl-OH (open circles). Vmax corresponds to 1600 nmol acetylene
formed per min per mg MoFe protein.
To distinguish whether or not propargyl-OH is only a flux
inhibitor (i.e. inhibitor of electron flow), or also a substrate that
competes with other substrates for available reducing equiv-
alents, we assayed for a product of propargyl-OH reduction
Ala
catalyzed by the α-70 MoFe protein. Gas chromatographic
analysis of the reaction vial headspace after incubation of
Ala
propargyl-OH with the α-70 MoFe protein under catalytic
conditions did not reveal the time-dependent accumulation of
any detectable propyne or allene, potential products of the
two-electron reduction of propargyl-OH. However, very small
amounts of the four-electron reduction product, propene, could
be detected. Nevertheless, the formation of propene occurred at
a rate of less than 5 nmol product per min per mg MoFe
protein, far below levels that could account for the propargyl-
OH inhibition of proton or N2 reduction catalyzed by the
Ala
α-70 MoFe protein.
The other remaining possible product from the two-electron
reduction of propargyl-OH is allyl alcohol (allyl-OH). As
described in the Experimental section, a GC-mass spectrometry
method was used for identification and quantitation of allyl-
Ala
OH as a product of α-70 MoFe protein catalyzed propargyl-
Fig. 3 Propargyl-OH inhibition of NH3 production (A) and H
2
OH reduction. When propargyl-OH was used as substrate for
Ala
production (B) catalyzed by the wild type (solid circles) and α-70
-
Ala
the α-70 MoFe protein a significant amount of allyl-OH
substituted (solid squares) MoFe proteins. NH formation assays were
3
formation could be detected. Under the conditions used for
conducted under 1 atm of N and H production assays were conducted
2
2
Ϫ1
these experiments a maximum level of 400 nmol min of
under 1 atm Ar. (N.B. 100% activity for wild type NH3 and H2
production were 1900 and 2300 nmol product per min per mg MoFe
Ala
allyl-OH was produced per mg of the α-70 MoFe protein.
Ala
protein, respectively. 100% activity for the α-70 -substituted MoFe
However, because of the inherent difficulties in accurately
quantifying allyl-OH, whether or not propargyl-OH is an
inhibitor of electron flux, in addition to its ability to serve as a
relatively effective substrate, could not be determined with
certainty. There was not a sufficient amount of allyl-OH form-
ation catalyzed by the wild type enzyme to permit its detection
by GC-mass spectrometry, consistent with the ineffectiveness of
propargyl-OH as an inhibitor of the wild type enzyme.
protein maximum for NH and H production were 800 and 1800 nmol
3
2
product per min per mg MoFe protein, respectively).
a potent inhibitor of both proton and N reduction catalyzed
2
Ala
by the α-70 MoFe protein. In each case, 2 mM propargyl-OH
resulted in approximately 50% inhibition. In contrast,
propargyl-OH is only a very poor inhibitor of the reduction of
the same substrates by the wild type MoFe protein. To evaluate
the capacity for propargyl-OH to inhibit acetylene reduction
Ala
Propyne is an effective substrate for the ꢀ-70 MoFe protein
Ala
catalyzed by the α-70 MoFe protein, a series of substrate
Ala
saturation curves were obtained using acetylene as the substrate
and increasing concentrations of propargyl-OH as the
inhibitor. The results of these measurements (Fig. 4) show that,
Because propargyl-OH proved to be a substrate for the α-70
MoFe protein, but not for the wild type protein, we determined
Ala
if propyne could also be used as a substrate for the α-70
Ala
for the α-70 MoFe protein, propargyl-OH is a reversible,
MoFe protein. In contrast to allyl-OH formation catalyzed by
propargyl-OH reduction, propene, the product of two-electron
reduction of propyne, can be easily and accurately determined
by gas chromatography. It is already known that propyne is an
competitive inhibitor of acetylene reduction having a K of
i
approximately 4.0 mM. In contrast, propargyl-OH is such a
poor inhibitor of acetylene reduction catalyzed by the wild type
22,23
enzyme that a K could not be accurately calculated.
extremely poor substrate for the wild type enzyme.
Data
i
8
04
J. Chem. Soc., Dalton Trans., 2002, 802–807

View Article Online
A. vinelandii acetylene-resistant strain is not significantly
affected in its ability to reduce N but is severely impaired in its
2
ability to reduce acetylene. The altered MoFe protein from this
Gly
strain was shown to have the α-subunit 69 residue substituted
Ser 20
Gly
by α-69
.
Within the wild type enzyme the α-69 residue
Val
is located immediately adjacent to the α-70 residue, whose
side-chain approaches one of the three geometrically identical
4
–6
4
α-96
Fe-4S faces of FeMo-cofactor
(Fig. 1). Another residue,
Arg
also approaches the same 4Fe-4S face of FeMo-
cofactor and potentially interacts with the α-69
Gly
residue
through hydrogen bonding of a nitrogen atom within its
Gly
guanidino group to the carbonyl oxygen of α-69 (Fig. 1).
Based on these observations we proposed that the 4Fe-4S face
Val
Arg
of FeMo-cofactor approached by α-70 and α-96 provides
Ser
an acetylene-binding site and that the α-69 substitution
Val
Arg
results in movement of either or both α-70 and α-96 so
that effective interaction of acetylene with the active site is
prevented.
Previous work has shown that, although acetylene is an
effective nitrogenase substrate, other alkynes such as propyne
23,26
are not.
An explanation offered for this observation is that
constraints within the substrate-binding pocket prevent the
26,27
binding of short chain alkynes larger than acetylene.
Evidence supporting this view was provided by the observation
that, while propyne is only a very poor nitrogenase substrate,
23
cyclopropene is a relatively effective substrate. Thus, if our
Val
hypothesis that the 4Fe-4S face approached by α-70 provides
the substrate-binding site is correct, we reasoned that shorten-
Val
ing the aliphatic side-chain of α-70 could result in expanding
the substrate reduction capability of nitrogenase to include
short chain alkynes in addition to acetylene. Results reported
Val
Ala
here show that substitution of the α-70 residue by α-70
does result in an altered MoFe protein that is able to catalyze
the effective reduction of propargyl-OH and propyne, neither
of which is an effective substrate for the wild type enzyme.
These results can be interpreted in light of what is known about
nitrogenase catalysis and the crystallographic structure of the
MoFe protein.
It is currently believed that during nitrogenase turnover elec-
trons are delivered from the P-cluster to the substrate reduction
site provided by FeMo-cofactor. Inspection of the MoFe
protein structure shows that a short helix spans the region
between the P-cluster and FeMo-cofactor (Fig. 1). This helix
Fig. 5 Propyne inhibition of H2 production (A) and reduction of
propyne to yield propene (B) catalyzed by the wild type (circles) and
Ala
α-70 -substituted MoFe protein (squares). 100% H2 production is
2
300 nmol per min per mg MoFe for wild type and 1800 nmol per min
Ala
per mg for the α-70 substituted MoFe protein. A Michaelis–Menten
kinetic fit for the reduction of propyne by the α-70 -MoFe protein
Cys
originates with α-62 , which provides a thiolate ligand to one
Ala
Val
P-cluster subfragment, and ends at α-70 capping one side of a
gives Km and Vmax values of 0.02 atm propyne and 1350 nmol propene
produced per min per mg MoFe protein, respectively.
specific 4Fe-4S face of FeMo-cofactor. Another short helix
Cys
extends from α-88 , which provides a bridging ligand to both
Arg
shown in Fig. 5 reveal that propyne is an effective inhibitor of
P-cluster subfragments, and ends at α-96 capping the other
Ala
proton reduction catalyzed by the α-70 MoFe protein, as
side of the same 4Fe-4S face of FeMo-cofactor. Thus, of the
three geometrically identical 4Fe-4S faces of FeMo-cofactor,
this one is in closest proximity to the P-cluster and, therefore,
most appropriately positioned for receiving electrons delivered
from the P-cluster. Hence, this 4Fe-4S face is an attractive
candidate for providing a substrate-binding site. Our data
provide support for this possibility because shortening of the
well as an effective substrate for that enzyme. The product of
Ala
propyne reduction catalyzed by the α-70 MoFe protein was
found to be propene on the basis of its gas chromatographic
retention time when compared to a propene standard. Data
shown in Fig. 5 also confirm that propyne is only a very poor
substrate for the wild type MoFe protein. In the case of pro-
Ala
Val
pyne reduction catalyzed by the α-70 MoFe protein, propane,
aliphatic side-chain of α-70 permits the accommodation of
the four-electron reduction product, could not be detected. It
was also found that, like the wild type MoFe protein, the α-70
MoFe protein does not catalyze the four-electron reduction of
acetylene to yield ethane.
larger alkyne substrates not effectively reduced by the wild type
enzyme. Although we cannot yet rule out the possibility
of indirect structural perturbations at another face of FeMo-
cofactor, or an indirect rearrangement in the environment
surrounding the Mo atom, we do not favor these possibilities
for two reasons. First, structural perturbations within the
FeMo-cofactor polypeptide environment are frequently mani-
fested by a significant alteration in the EPR spectrum, yet there
are no such alterations in the EPR spectrum in the altered
Ala
Discussion
An approach frequently used to identify active sites in enzymes
involves using genetic selection for strains resistant to the
physiological effects of substrate analogs. We previously used
this approach to isolate a mutant strain of A. vinelandii that is
Ala
MoFe protein having the α-70 substitution. Second, substitu-
tions that significantly alter the polypeptide environment of
20
resistant to the short chain alkyne acetylene. Acetylene is an
FeMo-cofactor would also be expected to have a severe effect
1
,2
Ala
effective nitrogenase substrate
and a potent inhibitor of
Nitrogenase isolated from an
on substrate reduction. However, the α-70
substitution
1
,24,25
physiological N reduction.
expands the ability of the altered MoFe protein to include
2
J. Chem. Soc., Dalton Trans., 2002, 802–807 805

View Article Online
propargyl-OH and propyne with only a modest effect on the
protein and 0.45 mg of Fe protein to give a 36 : 1 molar ratio of
ability of the enzyme to reduce protons, acetylene or N . Based
on these considerations we conclude that, for the α-70 MoFe
Fe protein to MoFe protein. Acetylene was freshly prepared for
each experiment by the reaction of calcium carbide and water.
Propyne (98% pure) was purchased from Aldrich. A propargyl-
OH stock solution was anaerobically prepared by the addition
of appropriate amounts of propargyl alcohol and 0.25 M
2
Ala
protein, propargyl-OH and propyne bind at the 4Fe-4S face
Val
Arg
capped by α-70 and α-96 . However, this conclusion does
not rule out the possibilities that either of the other two 4Fe-4S
faces of FeMo-cofactor, or the Mo atom, are able to provide
substrate-binding sites under certain conditions. These issues
will need to be resolved with other experiments. Nevertheless,
our working model is that Mo is not directly involved in provid-
ing a substrate-binding site and that most or all nitrogenase
substrates are bound and are reduced at the same 4Fe-4S face.
Inspection of the resting-state crystal structure of the MoFe
protein indicates that when van der Waals repulsion forces are
considered there is very little room for substrates other than
protons to have access to the 4Fe-4S face of FeMo-cofactor
Hepes
[N-(2-hydroxyethyl)piperazine-NЈ-(2-ethanesulfonic
acid)] buffer pH 7.4 to an evacuated crimp sealed vial to give
a final concentration of 1 M. Assays were initiated by the
addition of Fe protein and the reaction allowed to proceed for
8 min while shaking in a 30 ЊC water bath and subsequently
terminated by the addition of 250 µl of a 0.4 M EDTA solution.
H2 production was monitored by injection of 200 µl of the
reactions gas phase into a Shimadzu GC-14 gas chromatograph
equipped with a Supelco 80/100 molecular sieve 5A column and
a thermal conductivity detector. Ethylene and propene produc-
tion were monitored using a Hewlett-Packard 5890A gas
chromatograph equipped with an Al O capillary column and
Val
Arg
capped by the α-70 and α-96 side-chains. In fact, there does
not appear to be any way to accommodate a bridging mode for
substrates between Fe atoms in the resting state of the enzyme.
This binding configuration is often cited as a likely possibility
because of the cis stereospecificity for proton addition when
2
3
a flame ionization detector.
Inhibition patterns were evaluated by examination of
Lineweaver–Burk double reciprocal plots of the kinetic assay
data. Michaelis–Menten constants were derived by fitting data
1,28
C D2 is used as substrate.
2
Consequently, if our model is
correct and substrates do bind at this 4Fe-4S face, then the
to the following hyperbolic equation: ν = ([S]Vmax)/(K ϩ [S]).
m
Val
Arg
side-chains of either or both α-70 and α-96
must move
during catalysis to accommodate substrate binding. Close
examination of the crystallographic model indicates that it is
unlikely the α-70 side-chain would be capable of substantial
Extraction and quantification of propargyl alcohol and allyl
alcohol from incubation mixtures
Val
The reaction mixture was quenched by addition of EDTA as
described above and the liquid was transferred to a 2 ml glass
movement without a significant rearrangement in the poly-
peptide structure. In contrast, there does appear to be sufficient
vial. Analytes were extracted from the solution by submerging
Arg
space to accommodate movement of the α-96 side-chain. We
41,42
an SPME fiber into the solution for 20 min.
(N.B. A detailed
find this possibility attractive for three reasons. First, movement
protocol for this method is described in Supelco, Bulletin 901).
The fiber was then inserted into a Hewlett-Packard 5790 gas
chromatograph that was interfaced to a VG7070E-HF mass
spectrometer. The system was operated in the splitless injection
mode with a helium head pressure of 12 psi. The gas chromato-
graph injection port [220 ЊC], and the purge valve were opened
after 0.25 min. The SPME fiber was conditioned for an add-
itional 0.75 min in the hot injection port prior to the next analy-
sis. The column oven temperature was programmed from 45 ЊC
Arg
of the α-96 is consistent with a possible role for its side-chain
as a proton shuttle between the pool of water surrounding
homocitrate and the substrate-binding site. Second, com-
4
,5
parison of MoFe protein crystal structures from A. vinelandii,
29
6
Clostridium pasteuranium, Klebsiella pneumoniae as well as
the A. vinelandii ADPؒALF stabilized complex reveals this
30
4
side-chain occupies slightly different positions in each structure,
indicating the possibility for their dynamic movement during
catalysis. Finally, in other amino acid substitution experiments
Ϫ1
to 120 ЊC at 5 ЊC min . The mass spectrometer source was
Arg
we found that shortening of the α-96 side-chain permits the
operated in the electron impact mode at 70 eV at 200 ЊC. The
magnetic field was scanned from 10 to 200 amu in 0.5 seconds.
Allyl-OH and propargyl-OH eluted at 1.22 min and 1.30 min,
respectively. Quantification was performed by measuring the
area under the chromatographic peak in a mass chromatogram
of the base peak in the mass spectrum of each compound (m/z
binding of certain substrates to the MoFe protein in the resting
31
state.
In summary, results reported here indicate that short chain
alkynes can bind and be reduced at a specific 4Fe-4S face of
FeMo-cofactor. These results should be useful for future
modeling studies and also provide the basis for other amino
acid substitution experiments designed to determine exactly
where and how substrates interact with FeMo-cofactor.
5
7 for allyl-OH and m/z 55 for propargyl-OH). Although the
gas chromatographic peaks were not completely resolved,
quantification was possible because there was no significant ion
intensity for m/z 55 in the propargyl-OH alcohol spectrum, nor
of m/z 57 in the allyl-OH spectrum.
Experimental
Strain construction, cell growth, and purification
Acknowledgements
Val
A mutant strain of A. vinelandii having the α-70 residue
Ala
substituted by α-70
was constructed using site-directed
We thank Jason Christiansen and Kim Harich for help with
data acquisition and useful discussions.
mutagenesis and 3 gene replacement methods previously
2–35
described in detail.
Cells were grown and derepressed for nif
35
gene expression and harvested as described previously. Crude
extracts were prepared by the osmotic shock method and
purified as previously described. Quantitation of protein was
performed by a modified biuret method using bovine serum
albumin as the standard and protein purity assessed by SDS-
polyacrylamide gel electrophoresis. All protein manipulations
were kept anaerobic through the use of a Schlenk apparatus
References
35
1 M. J. Dilworth, Biochim. Biophys. Acta, 1966, 127, 285.
2 R. Schöllhorn and R. H. Burris, Proc. Natl. Acad. Sci. USA, 1967,
58, 213.
3 J. Christiansen, D. R. Dean and L. C. Seefeldt, Annu. Rev. Plant
Physiol. Plant Mol. Biol., 2001, 52, 269.
36
37
4
5
J. Kim and D. C. Rees, Nature, 1992, 360, 553.
J. W. Peters, M. H. B. Stowell, S. M. Soltis, M. G. Finnegan,
M. K. Johnson and D. C. Rees, Biochemistry, 1997, 36, 1181.
38
fitted with a BASF catalyst tower.
Kinetic assays
6 S. M. Mayer, D. M. Lawson, C. A. Gormal, S. M. Roe and
B. E. Smith, J. Mol. Biol., 1999, 292, 871.
The overall technique and reaction mixture composition are
7
V. K. Shah and W. J. Brill, Proc. Natl. Acad. Sci. USA, 1977, 74,
39,40
described elsewhere.
Each assay contained 0.05 mg of MoFe
3249.
8
06 J. Chem. Soc., Dalton Trans., 2002, 802–807

View Article Online
8
9
K. Fisher, W. E. Newton and D. J. Lowe, Biochemistry, 2001, 40,
333.
R. N. F. Thorneley and D. J. Lowe, Biochem. J., 1984, 224, 903.
28 R. W. F. Hardy, R. D. Holsten, E. K. Jackson and R. C. Burns, Plant
3
Physiol., 1968, 43, 1185.
29 J. T. Bolin, N. Campobasso, S. W. Muchmore, T. V. Morgan
and L. E. Mortenson, in Molybdenum enzymes, cofactors and
model systems, ed., E. I. Stiefel, D. Coucouvanis and W. E.
Newton, American Chemical Society, Washington, DC, 1993, pp.
186–195.
1
1
1
1
1
1
1
1
0 R. N. F. Thorneley and D. J. Lowe, Biochem. J., 1984, 224, 887.
1 D. J. Lowe and R. N. F. Thorneley, Biochem. J., 1984, 224, 895.
2 D. J. Lowe and R. N. F. Thorneley, Biochem. J., 1984, 224, 877.
3 H. B. Deng and R. Hoffmann, Angew. Chem., 1993, 32, 1062.
4 I. G. Dance, Aust. J. Chem., 1994, 47, 979.
30 H. Schindelin, C. Kisker, J. L. Sehlessman, J. B. Howard and
D. C. Rees, Nature, 1997, 387, 370.
31 P. M. C. Benton, S. M. Mayer, J. Shao, B. M. Hoffman, D. R. Dean
and L. C. Seefeldt, Biochemistry, 2001, in press.
32 W. J. Page and M. von Tigerstrom, J. Bacteriol., 1979, 139,
1058.
33 A. C. Robinson, B. K. Burgess and D. R. Dean, J. Bacteriol., 1986,
166, 180.
34 M. R. Jacobson, V. L. Cash, M. C. Weiss, N. F. Laird, W. E. Newton
and D. R. Dean, Mol. Gen. Genet., 1989, 219, 49.
35 J. Christiansen, P. J. Goodwin, W. N. Lanzilotta, L. C. Seefeldt and
D. R. Dean, Biochemistry, 1998, 37, 12611.
36 V. Chromy, J. Fischer and V. Kulhanek, Clin. Chem., 1974, 20, 1362.
37 U. K. Laemmlli, Nature, 1970, 227, 680.
38 B. K. Burgess, D. B. Jacobs and E. I. Stiefel, Biochem. Biophys. Acta,
1980, 614, 196.
39 J. W. Peters, K. Fisher and D. R. Dean, J. Biol. Chem., 1994, 269,
28076.
5 I. G. Dance, Chem. Commun., 1997, 2, 165.
6 D. Sellmann and J. Sutter, J. Biol. Inorg. Chem., 1996, 1, 587.
7 D. Sellmann and A. Fursattel, Coord. Chem. Rev., 2000, 200–202,
5
45.
1
1
2
2
8 K. L. C. Gronberg, C. A. Gormal, M. C. Durrant, B. E. Smith and
R. A. Henderson, J. Am. Chem. Soc., 1998, 120, 10613.
9 R. K. Szilagyi, D. G. Musaev and K. Morokuma, Inorg. Chem.,
2
001, 40, 766.
0 J. Christiansen, V. L. Cash, L. C. Seefeldt and D. R. Dean, J. Biol.
Chem., 2000, 275, 11459.
1 J. Christiansen, L. C. Seefeldt and D. R. Dean, J. Biol. Chem., 2000,
2
75, 36104.
2
2
2 R. W. F. Hardy and E. K. Jackson, Fed. Proc., 1967, 26, 725.
3 C. E. McKenna, M.-C. McKenna and C. W. Huang, Proc. Natl.
Acad. Sci. USA., 1979, 76, 4773.
2
4 J. C. Hwang, C. H. Chen and H. Burris, Biochim. Biophys. Acta,
1
973, 29, 256.
2
2
5 J. M. Rivera-Ortiz and R. H. Burris, J. Bacteriol., 1975, 123, 537.
6 C. E. McKenna, in Molybdenum and Molybdenum Containing
Enzymes, ed., M. Coughlan, Pergamon Press, New York, 1980.
7 R. C. Burns and R. W. F. Hardy, in Molecular Biology, Biochemistry
and Biophysics Series, ed., A. Kleinzeller, G. F. Springer and
H. G. Wittman, Springer-Verlag, New York, 1975, vol. 21.
40 C. H. Kim, W. E. Newton and D. R. Dean, Biochemistry, 1995, 34,
2798.
41 Supelco, Inc., Bulletin, Sigma-Aldrich Co., 1997, 901, 65.
42 D. A. Cassada, Y. Zhang, D. D. Snow and R. F. Spalding, Anal.
Chem., 2000, 72, 4654.
2
43 N. Geux and M. C. Peitsch, Electrophoresis, 1997, 18, 2714.
J. Chem. Soc., Dalton Trans., 2002, 802–807
807