catalytic cycle. However, there are also differences between the two schemes,
primarily in the stable sites of protonation. Figure 2a, the cycle including H sred ,
suggests that the organic ligand is protonated first. Then, following the second
reduction, the proton is ambiguously associated with the H-cluster until a second
protonation event. In contrast, Fig. 2b, depicting the cycle without H sred , suggests
that the first proton coordinates directly to the distal Fe.
2.1.2 [NiFe]-Hydrogenases
[NiFe]- and [FeFe]-hydrogenases are not evolutionarily related, but the need to
perform the same catalysis has resulted in closely related active sites in an example
of convergent evolution. Most obviously, the Fe in the [NiFe]-hydrogenase active site
is also coordinated by CO and CN
À ligands (Fig. 1b). Second, both sites are bimetallic. In addition, in both cases, the two metals are also bridged by thiolate ligands. In the
[NiFe] enzyme, these are provided by cysteine side chains. The distal iron of the
[FeFe]-hydrogenase can be thought of as substituted by the Ni of [NiFe]hydrogenases, but then the similarities start to break down. The primary coordination
sphere of the nickel is completed by two terminal cysteine thiolate ligands. Although
[NiFe]-hydrogenases do not feature a coordinated cubane cluster, it is worth noting
that all hydrogenases of this type include an [FeS] cluster, referred to as the proximal
cluster when other [FeS] clusters are present, near the [NiFe] active site. The [FeS]
clusters present in both [NiFe]- and [FeFe]-hydrogenases are essential for providing
an electrical linkage between the buried active site and the surface of the protein at
which the physiological partner can give or receive reducing equivalents. In some
[NiFe]-hydrogenases, the proximal cluster may also play a role in protecting the active
site from irreversible reactions with oxygen [18]. Finally, although extrinsic ligands
coordinate the H-cluster at a terminal position, spectroscopic evidence has shown that
many extrinsic ligands coordinate the [NiFe]-site in a bridging mode [19, 20].
Using primarily EPR signals associated with paramagnetic Ni states and the
FTIR signals associated with the vibrations of the diatomic ligands, a number of
different redox states of the [NiFe] active site have been identified [21]. Evidence
suggests that the Fe atom remains in a low spin Fe(II) state and all redox transitions
occur at the Ni ion which cycles between the Ni(II) and Ni(III) states. Under high
potential (oxidizing) or aerobic conditions, two different inactive Ni(III) states,
both spectroscopically and kinetically distinguishable, are formed. The Ni-A state
is reactivated on long timescales, whereas Ni-B requires shorter timescales to
reactivate. Crystal structures have suggested that both states contain a bridging
ligand derived from oxygen, but the chemical difference between these two states
remains unclear [20, 22]. Three catalytically competent states have also been
identified: Ni–Si a , Ni–C, and Ni–R. As shown in Fig. 3, the two most reduced,
Ni–C and Ni–R, are thought to contain a bridging hydride ligand.
Figure 3 shows a proposed catalytic mechanism for [NiFe]-hydrogenases. For
hydrogen oxidation, the catalytic cycle starts at the Ni–SI a state, a Ni(II) state with
no bridging ligand. Hydrogen binds the active site and is heterolytically cleaved to
produce the Ni–R state, a Ni(II) species with a bridging hydride and likely a
238
L. Gan et al.
primarily in the stable sites of protonation. Figure 2a, the cycle including H sred ,
suggests that the organic ligand is protonated first. Then, following the second
reduction, the proton is ambiguously associated with the H-cluster until a second
protonation event. In contrast, Fig. 2b, depicting the cycle without H sred , suggests
that the first proton coordinates directly to the distal Fe.
2.1.2 [NiFe]-Hydrogenases
[NiFe]- and [FeFe]-hydrogenases are not evolutionarily related, but the need to
perform the same catalysis has resulted in closely related active sites in an example
of convergent evolution. Most obviously, the Fe in the [NiFe]-hydrogenase active site
is also coordinated by CO and CN
À ligands (Fig. 1b). Second, both sites are bimetallic. In addition, in both cases, the two metals are also bridged by thiolate ligands. In the
[NiFe] enzyme, these are provided by cysteine side chains. The distal iron of the
[FeFe]-hydrogenase can be thought of as substituted by the Ni of [NiFe]hydrogenases, but then the similarities start to break down. The primary coordination
sphere of the nickel is completed by two terminal cysteine thiolate ligands. Although
[NiFe]-hydrogenases do not feature a coordinated cubane cluster, it is worth noting
that all hydrogenases of this type include an [FeS] cluster, referred to as the proximal
cluster when other [FeS] clusters are present, near the [NiFe] active site. The [FeS]
clusters present in both [NiFe]- and [FeFe]-hydrogenases are essential for providing
an electrical linkage between the buried active site and the surface of the protein at
which the physiological partner can give or receive reducing equivalents. In some
[NiFe]-hydrogenases, the proximal cluster may also play a role in protecting the active
site from irreversible reactions with oxygen [18]. Finally, although extrinsic ligands
coordinate the H-cluster at a terminal position, spectroscopic evidence has shown that
many extrinsic ligands coordinate the [NiFe]-site in a bridging mode [19, 20].
Using primarily EPR signals associated with paramagnetic Ni states and the
FTIR signals associated with the vibrations of the diatomic ligands, a number of
different redox states of the [NiFe] active site have been identified [21]. Evidence
suggests that the Fe atom remains in a low spin Fe(II) state and all redox transitions
occur at the Ni ion which cycles between the Ni(II) and Ni(III) states. Under high
potential (oxidizing) or aerobic conditions, two different inactive Ni(III) states,
both spectroscopically and kinetically distinguishable, are formed. The Ni-A state
is reactivated on long timescales, whereas Ni-B requires shorter timescales to
reactivate. Crystal structures have suggested that both states contain a bridging
ligand derived from oxygen, but the chemical difference between these two states
remains unclear [20, 22]. Three catalytically competent states have also been
identified: Ni–Si a , Ni–C, and Ni–R. As shown in Fig. 3, the two most reduced,
Ni–C and Ni–R, are thought to contain a bridging hydride ligand.
Figure 3 shows a proposed catalytic mechanism for [NiFe]-hydrogenases. For
hydrogen oxidation, the catalytic cycle starts at the Ni–SI a state, a Ni(II) state with
no bridging ligand. Hydrogen binds the active site and is heterolytically cleaved to
produce the Ni–R state, a Ni(II) species with a bridging hydride and likely a
238
L. Gan et al.