Communications
+
[5]
concomitant reduction of H to H . Thus, the hydrogen used
2
to generate the hydrocarbons could come either directly from
+
[1]
H
(analogous to the nitrogenase-based N reduction that
2
+
À
involves the addition of H and e to N ) or indirectly from H
2
2
(
analogous to the industrial, Fischer–Tropsch-based hydro-
[
6]
carbon formation that involves the hydrogenation of CO).
Previous work showed that the formation of hydrocarbons by
V nitrogenase was inhibited by the addition of increasing
amounts of H , suggesting that H was unlikely the hydrogen
2
2
source for this reaction. Here, we report the results of isotope
experiments for the V nitrogenase-catalyzed reduction of CO,
which include the identification of the hydrogen source, as
well as the detection of a new product.
[
5]
As reported earlier, when CO is reduced by V nitro-
Figure 3. Formation of propylene by V nitrogenase. a) Time-dependent
+
1
2
12
12
genase in the presence of H (i.e., a H O-based buffer), C H ,
formation of C D (*) and C H (*) in the presence of 100% CO.
2
2
4
3
6
3
6
C H , and C H can be detected by GC–MS at m/z ratios of
The samples were prepared in D
2
O (*)- and H O (*)-based buffers.
2
2
6
3
8
The data are presented as mean value Æ standard deviation (N=5).
2
8.032, 30.048, and 44.064, respectively (Figure 2a–c, experi-
1
2
12
13
+ +
b) GC–MS analysis of C H (top), C D (middle), and C D
3
6
3
6
3
6
ment 1). Upon substitution of H by D (i.e., the D O-based
2
(
bottom) formed by vanadium nitrogenase. The samples were pre-
buffer), the masses of these products shift by + 4, + 6, and + 8,
pared in H O- (top) or D O- (middle and bottom) based buffers and
2
2
respectively, consistent with the formation of C D , C D , and
12
13
2
4
2
6
contained 100% CO (top and middle) or 100% CO (bottom). The
C D (Figure 2a–c, experiment 2). Additional mass shifts of
3
8
products were traced at the following mass-to-charge (m/z) ratios:
1
3
12
12
13
+
2, + 2, and + 3 are observed when CO is supplied together
42.048: C H (top), 48.084: C D (middle), and 51.093: C3D6
3 6 3 6
+
with D , corresponding to the formation of double-labeled
products, C D , C D , and C D8 (Figure 2a–c, experi-
ment 3). Apart from the mass shifts, the incorporation of D
(bottom).
1
3
13
13
2
4
2
6
3
+
+
in these products is further demonstrated by the fact that they
elute slightly faster than their respective protonated counter-
parts (see Figure 2a–c, experiment 1 vs. experiment 2). Such a
behavior is characteristic of deuterated compounds, which
usually show a decrease in the retention time on the nonpolar
The identification of H ions as the hydrogen source for
hydrocarbon formation by V nitrogenase points to a parallel-
ism between the enzyme-based CO and N reduction, as both
2
+
À
reactions involve the ATP-dependent addition of H and e
to the substrate and the concomitant evolution of H as a side
2
[
7]
GC–MS columns. However, when 5.5% D2 is supplied
product (Figure 4). Such an analogy implies some mechanistic
+
together with H , no deuterated products can be detected
(
Figure 2a–c, experiment 4), although the formation of C H ,
2 4
C H , and C H is unaffected (Figure 2a–c, experiment 5).
2
6
3
8
The same effect is reproduced when 1.2% D is supplied to
2
the reaction (data not shown). Apparently, the hydrogen in
the hydrocarbon does not come from the concomitant
evolution of H2 by V nitrogenase, as 1.2 and 5.5% D2
Figure 4. Proposed reaction schemes for the reduction of CO (left)
represent the amounts of H produced at 10 min and 1 h,
2
[
5]
and N (right) by V nitrogenase. Both reactions involve the ATP-
2
respectively, concurrent with the hydrocarbons. Together,
these results firmly establish the soluble H ions (rather than
+
dependent protonation of substrates and the concomitant evolution of
H2.
H ) as the source of hydrogen for V-nitrogenase-based
2
hydrocarbon formation.
+
+
Interestingly, when H is replaced by D in the reaction
mixture, a new hydrocarbon product can be detected. The
time-dependent formation of this product is observed in the
similarities between the two reactions, particularly consider-
ing the isoelectronic properties of CO and N . On the other
2
hand, the reactions of CO and N reduction differ in that the
2
+
+
presence of D , but not in the presence of H (Figure 3a).
GC–MS analysis further confirms the identity of this product
as deuterated propylene (C D ), which displays an m/z ratio
former favors the reductive formation of CÀC bonds from CO
and the progressive extension of hydrocarbon chains, whereas
the latter supports the complete cleavage of the triple bond of
N and the formation of fully reduced NH . The deuterium
3
6
of 48.084 (Figure 3b, middle). There is a further shift in the
2
3
1
3
mass of this hydrocarbon product by + 3 when CO is used in
effect on the former reaction is especially interesting, as the
solvent isotope effects of D O/H O are well-documented and
+
combination with D as the substrates, consistent with the
2
2
1
3
formation of C D in this reaction (Figure 3b, bottom). In
can often be used to address the mechanistic questions of
3
6
+
[8]
contrast, C H is not detected by GC–MS when H is supplied
enzymatic reactions. In the current case, the deuterium-
3
6
to the reaction (Figure 3b, top). It is likely, therefore, that
C H is an intermediate that occurs during the V-nitrogenase-
dependent formation of C D could be explained by inverse
3
6
kinetic isotope effects (i.e., k /k < 1) that favor the forma-
3
6
H
9–11]
D
[
catalyzed extension of the hydrocarbon chain; however, it is
normally not detectable because of its rapid turnover to C H
tion of deuterated products.
However, such an effect is
not consistently observed in the V-nitrogenase-catalyzed
formation of other hydrocarbon products; rather, there is an
3
8
+
in the presence of H .
5
546
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 5545 –5547