Reduction of fluorinated cyclopropene by nitrogenase

Article abstract of DOI:10.1021/ja3121058

Reduction of the first known halogen-containing substrate by nitrogenase (N₂ase), 3,3-difluorocyclopropene (DFCP), was investigated. Reduction requires both N₂ase proteins (MoFe and Fe protein), ATP, and an exogenous reductant (dithionite, DT), as with N₂ and known alternative substrates of the enzyme. Two major products providing evidence for reductive C-F bond cleavage were confirmed, propene (P1, requiring 6e ^-/6H⁺) and 2-fluoropropene (P2, 4e^-/4H ⁺). Both were identified by GC-MS and NMR spectroscopy, and had the same K_m constants (0.022 atm, 5.4 mM). Reduction of 1,2-dideuterated DFCP (d₂-DFCP) further revealed that (i) in both P1 and P2, two deuterium atoms are retained, one on carbon-1 and one on carbon-3, indicating that C=C bond cleavage rather than C-C bond cleavage is involved during DFCP reduction at least to P2 (assuming no F migration); (ii) no selectivity was observed in formation of cis and trans isomers of 1,3-d₂-2- fluoropropene, whereas cis-1,3-d₂-propene is the predominant 1,3-d₂-propene product, indicating that one of the bound reduction intermediates on the pathway to propene is constrained geometrically. A reduction mechanism, consistent with hydride transfer as a key step, is discussed. Reductive C-F bond cleavage is an ability of N₂ase that further demonstrates the unique and remarkable scope of its catalytic prowess.

Full text of DOI:10.1021/ja3121058

Article
pubs.acs.org/JACS
Reduction of Fluorinated Cyclopropene by Nitrogenase
Feng Ni,† Chi Chung Lee,‡ Candy S. Hwang,† Yilin Hu,‡ Markus W. Ribbe,‡ and Charles E. McKenna*,†
†Department of Chemistry, University of Southern California, Los Angeles, California 90089-0744, United States
‡Department of Molecular Biology and Biochemistry, University of California, Irvine, California 92697-3900, United States
S
* Supporting Information
ABSTRACT: Reduction of the first known halogen-containing substrate by nitrogenase (N2ase), 3,3-difluorocyclopropene
(DFCP), was investigated. Reduction requires both N2ase proteins (MoFe and Fe protein), ATP, and an exogenous reductant
(dithionite, DT), as with N2 and known alternative substrates of the enzyme. Two major products providing evidence for
reductive C−F bond cleavage were confirmed, propene (P1, requiring 6e−/6H+) and 2-fluoropropene (P2, 4e−/4H+). Both were
identified by GC-MS and NMR spectroscopy, and had the same Km constants (0.022 atm, 5.4 mM). Reduction of 1,2-
dideuterated DFCP (d2-DFCP) further revealed that (i) in both P1 and P2, two deuterium atoms are retained, one on carbon-1
and one on carbon-3, indicating that CC bond cleavage rather than C−C bond cleavage is involved during DFCP reduction at
least to P2 (assuming no F migration); (ii) no selectivity was observed in formation of cis and trans isomers of 1,3-d2-2-
fluoropropene, whereas cis-1,3-d2-propene is the predominant 1,3-d2-propene product, indicating that one of the bound reduction
intermediates on the pathway to propene is constrained geometrically. A reduction mechanism, consistent with hydride transfer
as a key step, is discussed. Reductive C−F bond cleavage is an ability of N2ase that further demonstrates the unique and
remarkable scope of its catalytic prowess.
FeMoCo identified as the site of substrate binding and
reduction.2,5
1. INTRODUCTION
Nitrogenase (N2ase) reduction of N2 to ammonia is coupled to
the hydrolysis of 16 equiv of adenosine triphosphate (ATP) to
adenosine diphosphate/inorganic phosphate (ADP/Pi) and is
accompanied by the formation of one molecule of H2.1−4 Thus,
in contrast to the familiar biochemical process of oxidative
phosphorylation, this enzyme mediates a reductive dephos-
phorylation, consuming energy to promote catalysis despite the
favorable thermodynamics of the reaction at ambient temper-
ature. There are at least three major known classes of N2ases.
Each type is encoded by a unique set of genes, and each has a
different combination of bound metals (i.e., Fe/Mo, Fe/V, and
Fe/Fe). The most widely studied N2ase is the molybdenum-
containing enzyme, which consists of two component metal-
loproteins, the MoFeP protein (MoFeP) and the Fe protein
(FeP). FeP transfers an electron to MoFeP coupled with the
hydrolysis of two MgATP. MoFeP contains two different types
of unusual metallocenters, the FeMo cofactor (FeMoCo) and
the 8Fe7S P-cluster. Each electron-transfer step between FeP
and MoFeP involves an obligatory cycle of association and
dissociation of the protein complex, with dissociation proposed
to be the rate-determining step for the overall reaction, and
Besides reduction of the natural substrates, N2 and H+, wild-
type N2ase can reduce a wide range of small molecules
1
−
containing a triple bond (HCN, CH3NC, HN3/N3 , C2H2).
Very recently, our appreciation of the catalytic versatility of
N2ase has expanded to include the reduction of CO, long
recognized as a noncompetitive inhibitor of N2ase.6−8
However, even slightly larger homologues of terminal alkynes
are very poor substrates (e.g., propyne), and nonterminal
alkynes are not reduced at all (e.g., 2-butyne).9 Simple alkenes
(e.g., C2H4) are also virtually unreactive with wild-type N2ase,9
but the strained-ring CC hydrocarbon cyclopropene (Km =
0.1 mM) is a relatively good substrate (N2: Km = 0.1 mM) that
undergoes both reductive ring cleavage to propene and
reduction to an alkane (cyclopropane).10,11 Diazirine (Km
=
0.05−0.09 mM), containing the azo (−NN−) group in a
strained, three-membered ring, has a Km similar to that of N2
itself and is reduced by N2ase to methane, ammonia, and
methylamine, consistent with reductive cleavage of both the
Received: December 18, 2012
Published: July 8, 2013
© 2013 American Chemical Society
10346
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
Table 1. Nitrogenase-Catalyzed Reduction of 3,3-Difluorocyclopropene
c
products formed nmol/min/mg MoFeP)
ab
,
expt. no.
assay mixture
FeP (mg/vial)
MoFeP (mg/vial)
FeP/MoFeP
propene (P1)
2-fluoropropene (P2)
ratio P1/P2
1
2
3
4
5
6
7
8
9
10
complete
complete
complete
complete
complete
complete
− ATP
− DT
− FeP
− MoFeP
0.2
0.2
0.4
1.0
4.0
8.0
1.5
1.5
1.5
1.5
4
1:5
1:1
6.6
11.9
15.4
19.1
15.8
17.3
nd
2.9
4.3
4.9
5.9
4.5
5.1
nd
2.3
2.7
3.1
3.3
3.5
3.4
−
−
−
−
0.8
0.8
0.8
0.8
0.8
0.2
0.2
0.2
0.2
2:1
5:1
20:1
30:1
30:1
30:1
30:1
trace
nd
trace
nd
30:1
nd
nd
a
b
Initial partial pressure of DFCP 0.03 atm in 1 atm Ar, determined manometrically. Reaction quenched by 0.1 mL of 100 mM EDTA (pH 7.5) after
15 min. Determined by GC.
c
NN and C−N bonds.12 cis- and trans-Dimethyldiazene are
reduced to the same products, with decreasing effectiveness
(Km = 60 and 500−600 mM, respectively).12,13 Simeonov and
McKenna also showed that monomethyldiazene inhibits H2
evolution and C2H2 reduction, and detected at least one
reduction product, methylamine.14 Monomethyldiazene and
also diazene, both of which are unstable under assay conditions,
were subsequently shown to give rise to spectroscopically
detectable enzyme-bound intermediates.15
Thus, cyclopropene is the only alkene known that is an
effective substrate of wild-type N2ase. To examine the effect of
electron deficiency on the CC bond reactivity, 3,3-
difluorocyclopropene (DFCP) is of interest as a potential
N2ase substrate. The C−F bond is typically unreactive
chemically due to its large bond strength, and fluorine has
minimal steric impact as a substituent.16 N2ase-catalyzed
reduction of fluorinated substrates is also of interest in relation
to biological degradation of halogenated compounds17,18 and
organometallic activation of C−F bonds.19−21 Apart from
representing a novel class of substrate containing fluorine,
DFCP might provide a protein-bound reduction intermediate
retaining the 19F nucleus, and thus might be useful for nuclear
spin resonance-dependent spectroscopies such as EPR/
ENDOR and ESEEM, which are important tools3,4,22 to
elucidate N2ase structure and function.
gas phase by gas chromatography−mass spectrometry (GC-
MS) and 1H and 19F NMR confirmed the products as propene
and 2-fluoropropene, by comparison with spectra of authentic
samples. As shown in Figure S1, the product peaks P1 and P2
were detected at tretention = 1.66 min (m/z = 41 [M − H]+,
propene) and at tretention = 2.56 min (m/z = 59 [M − H]+,
consistent with a fluoropropene). The 1H NMR spectrum
(Figure S2) of the product mixture is consistent with a mixture
(∼1:1) of authentic propene and 2-fluoropropene. The 19F
NMR spectrum of the reduction product eluting at 2.56 min
(Figure S3) is identical to that of authentic 2-fluoropropene.
Pertinent data concerning the reduction are presented in Table
1. Neither product was generated by control mixtures lacking
FeP, MoFeP, ATP, or DT. The reduction requirements for
DFCP are therefore the same as those of other N2ase
substrates, including N2, acetylene, and cyclopropene. H2 (0.1
atm) could not be substituted for DT as a reductant. It is
important to note that the propene/2-fluoropropene ratio
varies when the ratio of FeP/MoFeP (electron flux) is changed.
Higher electron flux favors the formation of propene, the 6e−/
6H+ reduction product.
The release of the implied third reduction product, fluoride,
from DFCP under assay conditions was also determined. The
amount of fluoride formed corresponding to addition of
enzyme to the assay mixture correlates with propene and 2-
fluoropropene formed by N2ase-catalyzed DFCP reduction
(Figure S4).
Here, we report a mechanistic investigation of DFCP as the
first halogen-containing substrate of N2ase and confirm23 that
the enzyme catalyzes a remarkable reductive C−F bond
cleavage to give the 6e−/6H+ and 4e−/4H+ reduction products
propene (P1) and 2-fluoropropene (P2), respectively, con-
sistent with a hydride-transfer step and CC cleavage,
suggested by analysis of the reduction products from 1,2-
dideuterated DFCP (1,2-d2-DFCP). We also describe an
improved synthesis of DFCP that conveniently provides the
pure compound in gram quantities.
Reduction of DFCP to Propene and 2-Fluoropropene:
Kinetic Analysis. Gas-phase studies with commercially
obtained 2-fluoropropene (0.1 atm) established its stability
under assay conditions. No propene was observed when 2-
fluoropropene (0.1 atm) was reacted with the N2ase assay
mixture, verifying that propene produced during DFCP
reduction did not arise from the reduction of free 2-
fluoropropene by N2ase. In addition, 2-fluoropropene was not
detectably a product from propene and F− (released during the
reduction). This was confirmed by incubating propene with
N2ase under varying concentrations of F− (as NaF). No 2-
fluoropropene was observed, ruling out the possibility that the
F− ion participated in the reaction at the active site to form the
observed organofluorine product, although the radius of F−
(133 pm in NaF crystals) is similar to that of hydride (146 pm
in NaH crystals).24 Incubation of C2H2 or C2H4 with N2ase
under varying concentrations of F− also did not give detectable
vinyl fluoride, which further supports this conclusion.
2. RESULTS
Reduction of DFCP to Propene and 2-Fluoropropene:
Product Identification by GC-MS and NMR. Exposure of
DFCP (0.05 atm in Ar) to 1 mL of a N2ase assay mixture
containing ATP (5 μM), MgCl2 (5 μM), creatine phosphate
(CP) (50 μM), creatine phosphokinase (CPK) (50 units),
HEPES (50 μM, pH 7.5), and DT (50 μM) results in the
formation of significant quantitities of two products (P1 and
P2), as detected by gas chromatography (GC) analysis
(Porapak N column) of the gas phase. Further analysis of the
10347
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
Double-reciprocal Lineweaver−Burk plots of the product-
forming reaction (Figure 1) were constructed for each product,
MoFeP = 20:1), the propene peak (tretention = 1.67 min) gives
m/z = 43 (d2-propene, [M − H]+), and the corresponding 2-
fluoropropene peak (tretention = 2.55 min) also shows a 2 Da
increase (m/z = 61, [M − H]+), consistent with retention of
both deuterium atoms present in the substrate.
The gas-phase product mixture was further characterized by
1H NMR, 2H-decoupled 1H NMR, and 1H-decoupled 2H NMR
(Figure 3, FeP/MoFeP = 20:1). The two major products were
identified as mixtures of d2-propenes and d2-2-fluoropropenes.
In the 1H NMR spectrum, a doublet at 4.463 ppm (JH−F = 16.5
Hz) was assigned to the cis-CHD proton in 1,3-d2-2-
fluoropropene. A 0.016 ppm upfield shift was observed for this
doublet25 (geminal deuterium isotope effect) compared to the
corresponding chemical shift of the cis-CH2 protons in 2-
fluoropropene. Similarly, the doublet at 4.175 ppm (JH−F = 48.5
Hz) was assigned to the trans-CHD proton in 1,3-d2-2-
fluoropropene (0.03 ppm upfield shift relative to the trans
proton in the CH2 of 2-fluoropropene). The doublet of
triplets at ∼1.9 ppm (JH−F = 16 Hz, JH−D = 2.5 Hz, 0.016 ppm
upfield shift relative to undeuterated 2-fluoropropene) is
assigned to CH2D− in cis- or trans-1,3-d2-2-fluoropropene.
Figure 1. Double-reciprocal Lineweaver−Burk Km plot for the
formation of propene and 2-fluoropropene from DFCP reduction by
N2ase. Km for both products is the same (0.022 atm, 5.4 mM).
2
This signal appears as a doublet (JH−F = 16 Hz) in the H-
1
decoupled H NMR spectrum, confirming the presence of one
deuterium coupled to the protons. The 2H NMR (1H-
decoupled) spectrum shows a doublet (JD−F = 2.0 Hz) at
1.926 ppm, consistent with the presence of fluorine in a
CH2D−C(F) moiety. By integration of the CHD proton
peaks in the 1H NMR spectrum, the ratio of d2-2-
fluoropropenes was determined as 1:1 cis-1,3-d2-2-fluoroprope-
ne:trans-1,3-d2-2-fluoropropene.
P1 and P2. Both plots were linear over the range examined and
extrapolated to the same Km value, 0.022 atm, giving Vm values
of 16.4 and 5.1 nmol/(min·mg MoFeP), respectively. Using a
value of 0.25 M/atm for the solubility of DFCP in assay buffer
at 30 °C, the estimated Km for DFCP is 5.4 mM, 50−100 times
larger than for its cyclic analogues, cyclopropene and diazirine.
Selectivity in DFCP Reduction: Reduction of d2-DFCP
in H2O. In order to explore the mechanism of DFCP
reductions, 1,2-dideuterio-DFCP (d2-DFCP) was synthesized
(see characterization in Supporting Information). d2-DFCP was
incubated with N2ase assay mixtures having FeP/MoFeP ratios
of 20:1 and 1:5, and the gas-phase products were subjected to
GC-MS analysis. As shown in the MS spectrum (Figure 2, FeP/
The remaining peaks in the 1H NMR spectrum were
assigned to d2-propene as follows (Figure 3). The multiplet at
1.70−1.75 ppm (JH−H = 6.0 Hz, JH−D = 2.0 Hz, 0.016 ppm
upfield shift relative to undeuterated propene) belongs to a
CH2D− group. This signal appears as a doublet (JH−H = 6.0
2
1
2
Hz) in the H-decoupled H NMR spectrum. The H NMR
(1H-decoupled) spectrum shows a singlet at 1.744 ppm,
consistent with the assignment to a CH2D− group in d2-
propene. The doublet signal at δ = 4.919 ppm (JH−H = 10.0 Hz)
is assigned to the CHD proton in cis-1,3-d2-propene. A 0.018
ppm upfield shift is observed for this doublet (geminal
deuterium isotope effect) relative to the corresponding peak
2
in propene. The H NMR (1H-decoupled) spectrum shows a
singlet at 5.075 ppm, supporting the presence of a CHD
group in cis-1,3-d2-propene. The weak doublet signal at 5.012
ppm (JH−H = 17.0 Hz, 0.022 ppm upfield shift relative to
propene) is assigned to the CHD moiety in trans-1,3-d2-2-
fluoropropene. Resonances at 5.75−5.85 ppm are contributed
by the H on carbon-2 in both cis-1,3-d2-propene and trans-1,3-
d2-propene. The ratio of cis-1,3-d2-propene:trans-1,3-d2-propene
was determined as 14:1 by integration of the CHD proton
peaks in the 1H NMR spectrum. It should be noted that signals
from the geminal protons of CH2CH(CH3−xDx) were
observed as two doublets at 4.936 ppm (merged with the
proton signal of CHD in cis-1,3-d2-propene) and 5.022 ppm,
respectively. This result indicates the presence of a minor
amount of d1-propene as a mixture of propene isomer,
consistent with the observation of a characteristic mass peak
(m/z = 41, [M − D]+ ion) for d1-propene in Figure 2B. Based
on the NMR and GC data analysis, the constitution of the
identified product mixture from d2-DFCP reduction is
summarized in Scheme 1. Figure S5 shows the dependence
of the d2-DFCP reduction product distribution under different
Figure 2. GC-MS spectra of products from d2-DFCP reduction by
N2ase (FeP/MoFeP = 20:1). (A) GC chromatogram. (B) MS of P1,
identified as d2-propene ([M − H]+, m/z = 43). (C) MS of P2,
identified as d2-2-fluoropropene ([M − H]+, m/z = 61).
10348
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
Figure 3. NMR spectra (500 MHz; CDCl3/C2Cl4 = 1:1) of products of N2ase-catalyzed reduction of d2-DFCP (FeP/MoFeP = 20:1). (A) 1H NMR
2
1
1
2
spectrum. (B) H-decoupled H NMR spectrum. (C) H-decoupled H NMR spectrum.
lower affinity and lower N2ase activity than cyclopropene,
Scheme 1. d2-Propene and d2-2-Fluoropropene Isomers
Produced by N2ase-Catalyzed Reduction of d2-DFCP (FeP/
MoFeP = 20:1)
suggesting that a possible topological restriction is imposed on
the substituents of carbon-3 (C-3) of the cyclopropene ring.
Alternatively, the low affinity and low activity of DFCP might
also be due to the effect of the electronegative fluorine
substituents on alkene reactivity. Although reduction chemistry
of DFCP has not been previously reported, it is recognized that
fluorine substitution at the methylene carbon of cyclopropene
results in significant lengthening of the CC bond and
shortening of the C−C single bonds,26 where fluorine
substitution at C-3 in cyclopropene appears to actually stabilize
the three-membered ring.27 This will decrease the “alkyne-like”
property of the strained CC bond, making it more “alkene-
like” and thus less effective as a N2ase substrate.
electron flux conditions: higher electron flux favors a relatively
greater amount of cis-1,3-d2-propene.
With cyclopropene, only 2e− reduction products were found
(cyclopropane and propene). With fluorine substitution, the
reduction products pattern is changed to 4e−/6e− product, 2-
fluoropropene/propene. These products require an unprece-
dented reductive C−F bond cleavage by N2ase. The “missing”
1,1-difluorocyclopropane product (bp −16 °C)28 might be
explained by a destabilizing effect on the three-membered ring
of cyclopropane by fluorine substitution, unlike the stabilizing
effect of fluorine on unsaturated cyclopropene.27 Such an effect
could make the proposed cyclopropane intermediate 5 in
Scheme 2 prone to release strain energy through ring-opening
or by stabilizing the ring through fluoride release, without
formation of 1,1-difluorocyclopropane. Assuming that no
fluorine migration takes place during reduction, the fluorine
3. DISCUSSION
Mechanistic Implications. N2ase cleaves the strained
cyclopropene ring of DFCP, adding reductively 2H, mirroring
its reduction of cyclopropene itself to propene (2e−). However,
with DFCP, additional electrons are transferred before product
release, cleaving one (2e−) or both (4e−) C−F bonds, giving 2-
fluoropropene and propene, respectively. The choice of 3,3-
difluorinated cyclopropene (DFCP) rather than a 1- or 2-
fluorinated cyclopropene was based on the known fact that in
alkyne substrates, extension of the carbon chain length along
the −CC− axis decreases effectiveness as a N2ase substrate
(e.g., propyne, butyne). Nevertheless, the Km (5.4 mM) and Vm
(16.4 and 5.1 nmol/(min·mg MoFeP)) values of DFCP imply
10349
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
Scheme 2. Proposed Mechanism for Reduction of DFCP Catalyzed by N2ase
atom on the 2-fluoro product provides a mechanistic marker
indicating that DFCP is cleaved symmetrically with respect to
its CC bond.
the stereochemistry of N2ase reductions.11,29 Comparing the
product distributions at high electron flux (FeP/MoFeP =
20:1) and low (FeP/MoFeP = 1:5), more cis isomer was
observed for d2-propene at higher electron flux (Figure S5).
This trend was also observed in d2-acetylene reduction
catalyzed by N2ase, where higher electron flux favors formation
of cis-d2-ethylene.11
In the d2-DFCP reduction experiment, both MS (Figure 2B)
and NMR (Figure 3) data indicate that a d1-propene is an
apparently minor product of the reduction, most likely 3-d-
propene (Figure 3A,B). Chemical exchange of d2-DFCP with
H2O is unlikely under the assay conditions. An enzyme-
catalyzed exchange of d2-DFCP with a protiated species at the
MoFe protein active site, e.g., addition of metal−hydride to the
d2-DFCP CC bond, could be followed by elimination of a
deuterium atom to generate d1-DFCP, which could then be
reduced to the d1-propene.
A DFCP reduction mechanism catalyzed by N2ase that
accounts for the observed results and other available
information is proposed in Scheme 2. In this mechanism, we
focus on substrate transformation leading to product, but do
not speculate on how cluster metals are involved in the
reduction, or the source of protons. The symbol M thus refers
to either Fe or Mo. Based on an intermediate trapped during
propargyl alcohol reduction by a mutant, an intermediate
alkene bound side-on to a single Fe ion30 was proposed. Initial
binding of DFCP to FeMoCo is analogously proposed to give
intermediate 5 (Scheme 2). Hydrometalation and dimetalation
of the electron-rich CC bond in olefins is well-established
chemistry.31 Oxidative addition of the C−F bond by a
transition metal, especially Ni0, Pd0, and Pt0 complexes,21 is
also well known, and this reaction requires the metal to be in a
low oxidation state. In a recent report, when low-coordinate
Fe(II) complexes were reacted with fluoroolefins having both
CC and C−F moieties, initial addition by the Fe(II)−
hydride of the CC groups, rather than insertion into C−F
bond, was observed.32
d2-DFCP reduction exhibits two important features. First, the
two deuteriums are retained in the major products (propene
and 2-fluoropropene) and are found about equally on carbon-1
and carbon-3 in both, indicating that CC bond cleavage is
the main reaction pathway rather than initial ring C−C bond
cleavage, but via a 4e− process involving the elimination of F−.
N2ase-catalyzed reduction of cyclopropene in D2O to cyclo-
propane gave a cis-1,2-dideuterated product, also consistent
with symmetrical addition of hydrogen across the CC
bond.11 Reduction of cyclopropene gave 1,3-d2-propene as the
major propene product, with a small amount of the 2,3-d2
isomer.11
The second notable result is that the reduction product 1,3-
d2-2-fluoropropene is formed as equal amounts of the cis and
trans isomers, while cis-1,3-d2-propene is the major isomer in
the 1,3-d2-propene product. This suggests that the reduction
intermediate on the pathway to propene is effectively
constrained and that the formation of the products may
involve distinct pathways. It has been proposed that both
electron flux and steric constraints around the active site affect
In Path A, reduction of 5 by two electrons and two protons
(plausibly as hydride species) results in the C−C bond cleavage
and intermediate 2. After further reduction of 2 by one electron
and one proton along with cleavage of the M−C bond,
intermediate 3 is proposed as the precursor to the final
products 4a and 4b. Since departure of each fluorine on C-3 is
10350
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
equally likely upon β-elimination, the final step entails
formation of a CC bond with no stereoselectivity, resulting
in the observed equivalent amounts of the isomeric 1,3-d2-2-
fluoropropenes. (It is also interesting to consider the observed
selectivity in terms of theoretically postulated non-equivalent
H-transfer pathways.5b)
4. CONCLUSIONS
DFCP is the first known halogen-containing substrate of N2ase,
which reduces DFCP to two detected reduction products,
propene and 2-fluoropropene, identified by GC-MS and NMR
spectroscopy. Both propene and 2-fluoropropene have the
same Km constants (0.022 atm, 5.4 mM), indicating that they
are reduction products of the same substrate. The fluorine
atoms create an “electron sink” that converts the strained-ring
cycloalkene, normally a 2e− substrate, into a 4e− or 6e−
substrate with reductive cleavage of one or both C−F bonds
to eliminate F−. Formation of a product that retains one
fluorine atom (2-fluoropropene) implies a bound reduction
intermediate that retains a C−F group, thus making available a
potential new (19F) NMR or EPR/ENDOR/ESEEM probe22,36
for detecting active-site-bound species.
Analysis of reduction products from 1,2-d2-DFCP indicates
that one deuterium atom is found on carbon-1 and one on
carbon-3 of both propene and 2-fluoropropene, indicating that
initial CC bond cleavage rather than ring C−C bond
cleavage is the major reaction path during DFCP reduction, at
least to the latter product where the F atom defines the original
CF2 carbon unless F migration has occurred, consistent with
CC side-on binding of substrate to the metal cluster in the
active site. As proposed in Scheme 2, during reduction to
propene both F also depart directly and the CC bond in the
substrate is cleaved. The reduction product 1,3-d2-2-fluoropro-
pene consisted of equivalent amounts of cis and trans isomers,
whereas cis-1,3-d2-propene was the major isomer in 1,3-d2-
propene product, suggesting that the reduction intermediate on
the propene formation pathway is constrained.
The other pathway (Path B), leading to propene, is proposed
to involve initial loss of fluoride from intermediate 5, facilitated
by hydride attack at the CF2 carbon via an SN2 mechanism. The
intermediate 5 is proposed to undergo a facial-selective attack
by hydride on the front or back face of the constrained three-
membered ring. Since 5 covers the M site, a hydride transferred
from the FeMoCo would preferentially attack 5 on the “back-
face” (blue arrow), instead of the “front-face”(red arrow), of the
three-membered carbon ring in 5 (Scheme 2), thus
predominantly producing 6a over 6b. 6a and 6b are further
reduced to intermediates 8a and 8b, respectively. 8a and 8b
then undergo “anti-elimination” of fluoride ion to give 9a and
9b, respectively. The proposed pathway to propene formation
involves attack of hydride on intermediate 5, implying a metal
hydride species available to the bound DFCP. This is supported
by a recently suggested mechanism proposing a FeMoCo
hydride species.33,34
As depicted in Scheme 2, formation of intermediates 2 and 5
should be competitive. Propene as the dominant product over
2-fluoropropene might be explained on the basis that the
fluoride release step (Path B) might be a more energetically
favorable than the three-membered ring-opening step (Path A).
In summary, not only does ring-strain imposed on the
−CH2CH2− group transform it from a nonsubstrate (by
wild-type N2ase) to a fairly efficient 2e− substrate (cyclo-
propene), but the latter is further converted into a 6e− substrate
(DFCP, mimicking N2 in this respect) by incorporation of two
F atoms. This is not a binding effect because DFCP is more
poorly bound than cyclopropene itself, but rather a novel
“electron sink” effect from fluorine substitution, in which 2e−
(4) or 4e− (9) is consumed to cleave C−F bonds rather than
C−C bonds. The results support the earliest indications13,23
that hydride transfer precedes protonation.33,34
Reductive cleavage of the C−F bond in DFCP by N2ase
demonstrates the extraordinary catalytic versatility of this
enzyme beyond its natural role in biological N2 fixation6,7
and may be compared to ATP-dependent metalloenzyme
reductions in biological C−F bond degradation.17,18
5. EXPERIMENTAL SECTION
Nitrogenase. N2ase components were purified from continuously
cultured Azotobacter vinelandii OP. The MoFe protein (Av1, 24.5 mg/
mL) had a specific activity of 2000 (specific activity is defined as
nanomoles of C2H2 reduced per milligram of protein per minute), and
the Fe protein (Av2, 17 mg/mL) had a specific activity of 1900 nmol/
(mg·min).37
Assay Reagents. ATP Stock Solution. ATP, MgCl2·6H2O, CP,
and HEPES were dissolved in H2O and adjusted to pH 7.5 with 1 M
NaOH, and CPK was added. The final solution contained 10 mM
ATP, 10 mM MgCl2, 100 mM CP, 100 mM HEPES, and 100 units/
mL CPK.
Implications for Dehalogenase Mechanisms. It has
long been known that transition metal nucleophiles readily
displace fluoride from highly fluorinated arenes and alkenes to
afford metal−arene and metal−vinyl complexes, respectively.
These reactions are commonly viewed as simple nucleophilic
substitution reactions.19 Our results document a novel C−F
cleavage mediated by N2ase that may offer new insights into
biological C−F bond degradation.
Under aerobic conditions, compounds such as fluoroben-
zoate, fluorophenol, and fluorobenzene can be catabolized by
microbial enzymes via established aromatic hydrocarbon
pathways. However, under anaerobic conditions, little is
known about degradation of fluoroaromatic compounds.17
Interestingly, benzoyl-CoA reductase from Thauera aromatica
can reduce 2-, 3-, or 4-fluorobenzoyl-CoA.35 This enzyme is an
oxygen-sensitive iron−sulfur protein with a molecular mass of
160 kDa containing two separate [2Fe-2S] clusters and two
interacting [4Fe-4S] clusters in its four subunits. Benzoyl-CoA
reductase can also catalyze the ATP-dependent reduction of
hydroxylamine (Km = 0.15 mM) and azide. It has been
suggested that some of its properties resemble those of N2ase,
which similarly overcomes the high activation energy for
dinitrogen reduction by coupling electron transfer to the
hydrolysis of ATP.35
DT Stock Solution. Solid sodium dithionite was placed in a septum-
stoppered vial, which was pumped and filled with argon repeatedly.
Argon-flushed H2O was added with swirling, to give a 100 mM DT
solution.
Nitrogenase-Catalyzed Reduction Assays. (i). Kinetic Assays.
Several 9.3 mL septum-stoppered glass vaccine bottles were evacuated
to <20 μm Hg and then filled to 1 atm with a DFCP (partial pressure
0.0036−0.05 atm)/argon gas mixture using a vacuum line manifold.
C2H6 (20 μL) (Matheson, CP grade) was injected into each bottle as
an internal GC standard. GC analysis aliquots (20 μL) were removed
from each bottle and replaced immediately by equivalent volumes of
argon. To each bottle mounted in a 30 °C shaker bath were added in
rapid sequence 0.5 mL of degassed (argon) ATP stock solution (5
μmol of ATP, 5 μmol of MgCl2, 50 μmol of CP, 50 μmol of HEPES,
and 50 units of CPK), 0.2 mL of DT stock solution (20 μmol of DT),
0.233 mL of degassed (argon) H2O, 0.059 mL of FeP, and 0.008 mL
of MoFeP (FeP:MoFeP is 20:1) to a final assay solution volume of 1.0
10351
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352

Journal of the American Chemical Society
Article
mL. After 20 min, 0.1 mL of 100 mM ethylenediaminetetraacetic acid
(EDTA) (pH 7.5) was injected into the assay vial to quench the
reduction. Gas aliquots (20 μL) were removed for GC analysis.
(ii). Reduction Time Course Assay. A series of assay bottles
containing 1 atm of a DFCP (0.03 atm)/argon gas mixture were
prepared, and reductions were initiated as described in (i). The
reactions were quenched by 0.1 mL of 100 mM EDTA (pH 7.5) at
time intervals of 0, 6, 20, 30, 45, and 60 min. Gas aliquots (20 μL)
were removed for GC analysis.
(iii). Reductions with Varied Electron Flux. A series of assay bottles
containing 1 atm of a DFCP (0.03 atm)/argon gas mixture were
prepared, and reductions were initiated as described in (i), except that
FeP:MoFeP ratios were 1:5, 1:1, 2:1, 5:1, 20:1, and 30:1 (quantities of
FeP and MoFeP are shown in Table 1). The reactions were quenched
after 15 min, and gas aliquots (20 μL) were removed for GC analysis.
Binding and Reduction at the Active Site. In Nitrogen Fixation:
Fundamentals and Applications; Tikhonovich, I. A., Provorov, N.
A.Romanov, V. I.Newton, W. E., Eds.; Kluwer Academic Publishers:
Dordrecht, 1995.
(14) Simeonov, A. M. Ph.D. Dissertation, University of Southern
California, 1998; p 242.
(15) (a) Barney, B. M.; Yang, T.-C.; Igarashi, R. Y.; Dos Santos, P.
C.; Laryukhin, M.; Lee, H.-I.; Hoffman, B. M.; Dean, , D. R.; Seefeldt,
L. C. J. Am. Chem. Soc. 2005, 127, 14960−14961. (b) Barney, B. M.;
Lukoyanov, D.; Yang, T.-C.; Dean, D. R.; Hoffman, B. M.; Seefeldt, L.
C. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 17113−17118.
(c) Lukoyanov, D.; Dikanov, S. A.; Yang, Z. Y.; Barney, B. M.;
Samoilova, R. I.; Narasimhulu, K. V.; Dean, D. R.; Seefeldt, L. C.;
Hoffman, B. M. J. Am. Chem. Soc. 2011, 133, 11655−11664.
(16) (a) Berkowitz, D. B.; Karukurichi, K. R.; de la Salud-Bea, R.;
Nelson, D. L.; McCune, C. D. J. Fluorine Chem. 2008, 129, 731−742.
(b) Yamazaki, T.; Taguchi, T.; Ojima, I. Unique Properties of Fluorine
and their Relevance to Medicinal Chemistry and Chemical Biology. In
Fluorine in Medicinal Chemistry and Chemical Biology; Oijima, I., Ed.;
John Wiley & Sons: Chichester, 2009; pp 1−46.
(17) Murphy, C. D. Biotechnol. Lett. 2010, 32, 351−359.
(18) Natarajan, R.; Azerad, R.; Badet, B.; Copin, E. J. Fluorine Chem.
2005, 126, 425−436.
(19) Kiplinger, J. L.; Richmond, T. G.; Osterberg, C. E. Chem. Rev.
1994, 94, 373−431.
(20) Amii, H.; Uneyama, K. Chem. Rev. 2009, 109, 2119−2183.
(21) Clot, E.; Eisenstein, O.; Jasim, N.; Macgregor, S. A.; McGrady, J.
E.; Perutz, R. N. Acc. Chem. Res. 2011, 44, 333−348.
(22) Lukoyanov, D.; Yang, Z. Y.; Barney, B. M.; Dean, D. R.;
Seefeldt, L. C.; Hoffman, B. M. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
5583−5587.
(23) McKenna, C. E.; Eran, H.; Nakajima, T.; Osumi, A. Curr. Persp.
Nitrogen Fixation 1981, 358.
(24) Wells, A. F. Structural Inorganic Chemistry; Clarendon Press:
ASSOCIATED CONTENT
* Supporting Information
■
S
Sample preparation and methods for analysis of reduction
products; synthesis and characterization of DFCP and d2-
DFCP. This material is available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
mckenna@usc.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by funding from the Dana and David
Dornsife College of the University of Southern California
(C.E.M.) and in part by a National Science Foundation
Graduate Research Fellowship under Grant No. DGE-0937362
(C.S.H.). We thank Inah Kang for assistance in preparing the
manuscript.
Oxford, 1984.
(25) Baird, M. C. J. Magn. Reson. 1974, 14, 117−120.
(26) Dolbier, W. R. J.; Battiste, M. A. Chem. Rev. 2003, 103, 1071−
1098.
(27) Wiberg, K. B.; Marquez, M. J. Am. Chem. Soc. 1998, 120, 2932−
2938.
(28) McBee, E. T.; Roberts, C. W.; Judd, G. F.; Chao, T. S. J. Am.
Chem. Soc. 1955, 77, 1292−1293.
REFERENCES
■
(1) In Catalysts for Nitrogen Fixation: Nitrogenases, Relevant Chemical
Models and Commercial Processes; Smith, B., Richards, R. L., Newton,
W. E., Eds.; Kluwer Academic Publishers: Dordrecht, 2004; p 1−340.
(2) Seefeldt, L. C.; Hoffman, B. M.; Dean, D. R. Annu. Rev. Biochem.
2009, 78, 701−722.
(3) Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Acc. Chem. Res.
2009, 42, 609−619.
(4) Hoffman, B. M.; Lukoyanov, D.; Dean, D. R.; Seefeldt, L. C. Acc.
Chem. Res. 2013, 46, 587−595.
(29) Benton, P. M.; Christiansen, J.; Dean, D. R.; Seefeldt, L. C. J.
Am. Chem. Soc. 2001, 123, 1822−1827.
(30) Lee, H. I.; Igarashi, R. Y.; Laryukhin, M.; Doan, P. E.; Dos
Santos, P. C.; Dean, D. R.; Seefeldt, L. C.; Hoffman, B. M. J. Am.
Chem. Soc. 2004, 126, 9563−9569.
(31) Rubin, M.; Rubina, M.; Gevorgyan, V. Chem. Rev. 2007, 107,
3117−3179.
(32) Vela, J.; Smith, J. M.; Yu, Y.; Ketterer, N. A.; Flaschenriem, C. J.;
Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc. 2005, 127, 7857−
7870.
(5) (a) Thorneley, R. N. F.; Lowe, D. J. In Molybdenum enzymes;
Spiro, T. G., Ed.; John Wiley and Sons: Chichester, 1985; pp 221−
284. (b) Dance, I. J. Am. Chem. Soc. 2007, 129, 1076−1088.
(6) Lee, C. C.; Hu, Y.; Ribbe, M. W. Science 2010, 329, 642.
(7) Hu, Y.; Lee, C. C.; Ribbe, M. W. Science 2011, 333, 753−755.
(8) Yang, Z. Y.; Dean, D. R.; Seefeldt, L. C. J. Biol. Chem. 2011, 286,
19417−19421.
(33) Seefeldt, L. C.; Hoffman, B. M.; Dean, D. R. Curr. Opin. Chem.
Biol. 2012, 16, 19−25.
(34) Lukoyanov, D.; Yang, Z. Y.; Dean, D. R.; Seefeldt, L. C.;
Hoffman, B. M. J. Am. Chem. Soc. 2010, 132, 2526−2527.
(35) Boll, M.; Fuchs, G. Eur. J. Biochem. 1995, 234, 921−933.
(36) Smoukov, S. K.; Kopp, D. A.; Valentine, A. M.; Davydov, R.;
Lippard, S. J.; Hoffman, B. M. J. Am. Chem. Soc. 2002, 124, 2657−
2663.
(9) Hardy, R. W. F.; Burns, R. C. Annu. Rev. Biochem. 1968, 37, 331−
358.
(10) McKenna, C. E.; McKenna, M. C.; Higa, M. T. J. Am. Chem. Soc.
1976, 98, 4657−4659. McKenna, C. E.; Huang, C. W. Nature 1979,
280, 609−611. McKenna, C. E.; Jones, J. B.; Eran, H.; Huang, C. W.
Nature 1979, 280, 611−612.
(11) McKenna, C. E.; McKenna, M. C.; Huang, C. W. Proc. Natl.
Acad. Sci. U.S.A. 1979, 76, 4773−4777.
(37) Wiig, J. A.; Lee, C. C.; Fay, A. W.; Hu, Y. L.; Ribbe, M. W.
Methods Mol. Biol. 2011, 766, 93−103.
(12) McKenna, C. E.; Simeonov, A. M.; Eran, H.; Bravo-
Leerabhandh, M. Biochemistry 1996, 35, 4502−4514.
(13) McKenna, C. E.; Simeonov, A. M. Dimethyldiazenes are
Reduced by Nitrogenase: Implications for the Mechanism of Substrate
10352
dx.doi.org/10.1021/ja3121058 | J. Am. Chem. Soc. 2013, 135, 10346−10352