was then incubated with apo-NifDK, and then re-isolated. EPR analysis of this
species not only revealed the retainment of the M-cluster, it also showed the
alteration of the line-shape and amplitude of the M-cluster specific signal and thus
suggested a change of the cofactor environment. This observation also implies that
another conformational change occurs as NifEN
lock−M interacts with apo-NifDK, as
confirmed by a Fe-chelation experiment which shows the M-cluster gained more
solvent exposure after NifEN
lock−M was incubated with apo-NifDK [77]. These
findings highlight the dynamic nature of the NifEN protein, which initially presents
L-cluster on the surface of the protein to allow maturation, then undergoes a first
conformational change to embed the newly generated M-cluster, and finally, upon
interaction with NifDK, performs a second structural rearrangement to push the
M-cluster outward in order to accommodate the transfer of the cofactor (Fig. 5a).
Analogous to NifEN, apo-NifDK undergoes a conformational change of its own
when it receives the M-cluster from NifEN. This was demonstrated by SAXS studies
which revealed a more compact structure of holo-NifDK compared to apo-NifDK
[50]. Together with the structural analysis of apo-NifDK, this indicates that as
M-cluster travels down a positively charged insertion funnel at NifDK surface, the
protein folds-inward to bury the cofactor within the matrix of the a-subunit [23]. The
pathway of M-cluster insertion can be mapped to three specific regions of NifDK:
(i) a ‘lid loop’ region, consisting of the residues a353 to a364; (ii) a ‘His triad’ region
which is formed by Hisa274, Hisa442, and Hisa451; and (iii) the ‘switch/lock’
region composed of Hisa442 and Trpa444 (Fig. 5b) [78–80]. The positively
charged residues in the ‘lip/loop’ region, in combination with the potential transient
ligand Hisa362, likely helps to transiently bind and guide the anionic M-cluster to
correctly reach its binding site. Subsequently, the ‘His triad’ is proposed to be a
transient binding site(s) for the M-cluster as it passes along the insertion path.
Finally, the ‘switch/lock’-residue Hisa442 binds the cofactor within the active site
and then switches position with Trpa444, of which the bulky side chain serves as a
lock that fastens down the cofactor (Fig. 5b). Accordingly, the mutation of each of
these residues above leads to NifDK variants with impaired M-cluster uptake and
retention abilities [78–80]. The effect of Hisa442 and Trpa444 as the
M-cluster-switch/lock have also been demonstrated on NifEN
lock , and it can be
anticipated that further engineering using the other insertion funnel residues might
eventually lead to artificial proteins that are efficient in M-cluster uptake.
4 Mechanism of Nitrogenase
4.1 Overview: Reactive Properties and the Thornley-Lowe
Cycle
The basic mechanistic properties of Mo-nitrogenase is best outlined by the
Lowe-Thorneley (LT) model (Fig. 6a) [81]. This model, which summarized a large
body of early work on nitrogenase kinetics, was schematized by David Lowe and
170
C.-C. Lee et al.
species not only revealed the retainment of the M-cluster, it also showed the
alteration of the line-shape and amplitude of the M-cluster specific signal and thus
suggested a change of the cofactor environment. This observation also implies that
another conformational change occurs as NifEN
lock−M interacts with apo-NifDK, as
confirmed by a Fe-chelation experiment which shows the M-cluster gained more
solvent exposure after NifEN
lock−M was incubated with apo-NifDK [77]. These
findings highlight the dynamic nature of the NifEN protein, which initially presents
L-cluster on the surface of the protein to allow maturation, then undergoes a first
conformational change to embed the newly generated M-cluster, and finally, upon
interaction with NifDK, performs a second structural rearrangement to push the
M-cluster outward in order to accommodate the transfer of the cofactor (Fig. 5a).
Analogous to NifEN, apo-NifDK undergoes a conformational change of its own
when it receives the M-cluster from NifEN. This was demonstrated by SAXS studies
which revealed a more compact structure of holo-NifDK compared to apo-NifDK
[50]. Together with the structural analysis of apo-NifDK, this indicates that as
M-cluster travels down a positively charged insertion funnel at NifDK surface, the
protein folds-inward to bury the cofactor within the matrix of the a-subunit [23]. The
pathway of M-cluster insertion can be mapped to three specific regions of NifDK:
(i) a ‘lid loop’ region, consisting of the residues a353 to a364; (ii) a ‘His triad’ region
which is formed by Hisa274, Hisa442, and Hisa451; and (iii) the ‘switch/lock’
region composed of Hisa442 and Trpa444 (Fig. 5b) [78–80]. The positively
charged residues in the ‘lip/loop’ region, in combination with the potential transient
ligand Hisa362, likely helps to transiently bind and guide the anionic M-cluster to
correctly reach its binding site. Subsequently, the ‘His triad’ is proposed to be a
transient binding site(s) for the M-cluster as it passes along the insertion path.
Finally, the ‘switch/lock’-residue Hisa442 binds the cofactor within the active site
and then switches position with Trpa444, of which the bulky side chain serves as a
lock that fastens down the cofactor (Fig. 5b). Accordingly, the mutation of each of
these residues above leads to NifDK variants with impaired M-cluster uptake and
retention abilities [78–80]. The effect of Hisa442 and Trpa444 as the
M-cluster-switch/lock have also been demonstrated on NifEN
lock , and it can be
anticipated that further engineering using the other insertion funnel residues might
eventually lead to artificial proteins that are efficient in M-cluster uptake.
4 Mechanism of Nitrogenase
4.1 Overview: Reactive Properties and the Thornley-Lowe
Cycle
The basic mechanistic properties of Mo-nitrogenase is best outlined by the
Lowe-Thorneley (LT) model (Fig. 6a) [81]. This model, which summarized a large
body of early work on nitrogenase kinetics, was schematized by David Lowe and
170
C.-C. Lee et al.