Anisotropic Magnetic Spin Interactions of Transition Metal …
51
4.3 Transition Metal Containing Enzymes
4.3.1 The g-Tensor Orientation in a Protein Single Crystal
The nickel atom of [NiFe]-hydrogenase active site gives rise to rhombic EPR spectra
from an S ½ spin system and reveal no hyperfine interaction (the hyperfine interaction is smaller than the EPR linewidth). Hydrogenase nickel EPR spectra were known
in the literature for a very long time but the discovery of a second transition metal
in the active site in the first protein crystal structure was not expected. The second
metal was later identified by
57 Fe Mössbauer experiments to be a non-redox active
iron atom. The ‘Ni–C’ EPR signal refers to a catalytic intermediate in the heterolytic
splitting of H 2 . Ni–C is light-sensitive and converted to the light-induced Ni–L state
at low temperature (below 100 K). Upon illumination, the EPR spectrum converts
from that of a Ni–C (g i 2.20, 2.14, 2.01) to that of a Ni–L species (2.30, 2.12,
2.05). Recently, Ni–L was discussed to be also involved in the catalytic mechanism.
The Ni–C and Ni–L states of the [NiFe]-hydrogenase from D. vulgaris Miyazaki
F were generated in situ in protein single crystals by incubation with H 2 . An analysis
of the orientation-dependent EPR spectra yielded the full g-tensors and their orientations in the crystal axes system for both Ni–C and Ni–L forms simultaneously (see
Fig. 6).
Fitting of the eight EPR transitions (see Fig. 6) from 2 species in the crystal with
four molecules per unit cell each, the assignment of the magnetic g-tensor principal
axes to the enzyme active site structure is not unambiguous.
The geometry of the catalytic center and the g-tensor orientations are closely
related since the interaction of the unpaired electron spin at the nickel is mediated
by the coordinating ligands and the crystal field thus determines the orientation of
the magnetic axes. In the oxidized forms, the coordination of the Ni atom in the
active site can be described as distorted octahedron with one empty ligand position.
In absence of a protein structure of the reduced enzyme at that time, an active site
geometry close to the oxidized forms was assumed. The shift of the smallest g-tensor
component from g 3 ≈ 2.01 to g 3 ≈ 2.05 is indicative of a change in redox state of
the nickel atom. Various candidates for models of the Ni–C and Ni–L forms were
investigated by computations. The calculations reproduce well the upshift of g 3 to
2.05 upon photoreduction from Ni(III) to Ni(I) when dissociation of the bridging
hydride is assumed. This detailed insight is not available from protein crystallography. The calculated g-tensor principal values from ZORA DFT calculations are in
good agreement with experiment (see Table 3) and the deviation from experiment is
in the typical range as was also found to the Ni(mnt)
−
2 model complex (see above).
DFT calculations of the orientation of the g-tensor principal axes in the molecular
structure of the active site helped to assign the paramagnetic sites in the crystal (see
Table 3). The calculated g-tensor orientation is in excellent agreement with only one
possible experimental assignment and enabled the analysis and interpretation of the
electronic structure of the Ni–C and Ni–L states simultaneously. It was suggested
on the basis of computational work that Ni–L originates from the Ni(III)-μH
− -
51
4.3 Transition Metal Containing Enzymes
4.3.1 The g-Tensor Orientation in a Protein Single Crystal
The nickel atom of [NiFe]-hydrogenase active site gives rise to rhombic EPR spectra
from an S ½ spin system and reveal no hyperfine interaction (the hyperfine interaction is smaller than the EPR linewidth). Hydrogenase nickel EPR spectra were known
in the literature for a very long time but the discovery of a second transition metal
in the active site in the first protein crystal structure was not expected. The second
metal was later identified by
57 Fe Mössbauer experiments to be a non-redox active
iron atom. The ‘Ni–C’ EPR signal refers to a catalytic intermediate in the heterolytic
splitting of H 2 . Ni–C is light-sensitive and converted to the light-induced Ni–L state
at low temperature (below 100 K). Upon illumination, the EPR spectrum converts
from that of a Ni–C (g i 2.20, 2.14, 2.01) to that of a Ni–L species (2.30, 2.12,
2.05). Recently, Ni–L was discussed to be also involved in the catalytic mechanism.
The Ni–C and Ni–L states of the [NiFe]-hydrogenase from D. vulgaris Miyazaki
F were generated in situ in protein single crystals by incubation with H 2 . An analysis
of the orientation-dependent EPR spectra yielded the full g-tensors and their orientations in the crystal axes system for both Ni–C and Ni–L forms simultaneously (see
Fig. 6).
Fitting of the eight EPR transitions (see Fig. 6) from 2 species in the crystal with
four molecules per unit cell each, the assignment of the magnetic g-tensor principal
axes to the enzyme active site structure is not unambiguous.
The geometry of the catalytic center and the g-tensor orientations are closely
related since the interaction of the unpaired electron spin at the nickel is mediated
by the coordinating ligands and the crystal field thus determines the orientation of
the magnetic axes. In the oxidized forms, the coordination of the Ni atom in the
active site can be described as distorted octahedron with one empty ligand position.
In absence of a protein structure of the reduced enzyme at that time, an active site
geometry close to the oxidized forms was assumed. The shift of the smallest g-tensor
component from g 3 ≈ 2.01 to g 3 ≈ 2.05 is indicative of a change in redox state of
the nickel atom. Various candidates for models of the Ni–C and Ni–L forms were
investigated by computations. The calculations reproduce well the upshift of g 3 to
2.05 upon photoreduction from Ni(III) to Ni(I) when dissociation of the bridging
hydride is assumed. This detailed insight is not available from protein crystallography. The calculated g-tensor principal values from ZORA DFT calculations are in
good agreement with experiment (see Table 3) and the deviation from experiment is
in the typical range as was also found to the Ni(mnt)
−
2 model complex (see above).
DFT calculations of the orientation of the g-tensor principal axes in the molecular
structure of the active site helped to assign the paramagnetic sites in the crystal (see
Table 3). The calculated g-tensor orientation is in excellent agreement with only one
possible experimental assignment and enabled the analysis and interpretation of the
electronic structure of the Ni–C and Ni–L states simultaneously. It was suggested
on the basis of computational work that Ni–L originates from the Ni(III)-μH
− -