CYTOCHROME c OX IDASE 445
X - ray crystallographic data (Fe – Cu = 3.916 Å ), and both being ∼ 0.3 Å longer
than that observed for the [( 82 ) III−−(^2 −+) II( )] F TPP Fe (EXAFS data O Cu TMPA
for solid and in solution). The authors believe that a combination of steric and
electronic factors is responsible for the differences observed. In comparing
their results with those of the CcO enzyme itself, the reference 155 authors
note that the peroxo linkage probably does not form in the enzyme because
the Fe – Cu distance is too long (4.4 – 5.3 Å ), although one X - ray structure shows
a peroxo ligand (PDB: 2OCC). Differences between the model compounds
discussed here and CcO protein are (1) the presence of a proximal (opposite
the O 2 - binding site) axial histidine ligand on heme a 3 in CcO, (2) the tyrosine
residue covalently bound to a Cu B histidine ligand (which may donate an H +
during the catalytic cycle), and (3) the coordination environment of the copper
center that may cause electronic structural differences and thus differences in
reactivity in the FeIII−−O 22 − CuII center. In spite of these shortcomings, the
models do provide a good starting point for understanding the activation and
cleavage of the dioxygen bond in cytochrome c oxidase.
Biomemetic efforts have also focused on the oxidized or “ resting state ” of
the heme a 3 – Cu B binuclear center in which the porphyrinate iron ion is in the
Fe(III) oxidation state and the Cu B ion exists as Cu(II). For instance, Fe III – X –
Cu II synthetic model compounds, where X = oxo - or hydroxo - bridged
complexes, could represent the CcO “ resting state ” or catalytic turnover inter-
mediates after dioxygen O – O bond cleavage. Cyanide - bridged compounds
could yield information on the toxicity of the CN − ligand. X - ray crystallo-
graphic studies on native CcO enzymes over the last 10 years have proposed
various connecting ligands for the oxidized enzyme ’ s heme a 3 – Cu B binuclear
center. At least one of these — PDB: 2OCC for the oxidized “ resting state ”
bovine heart enzyme — proposed a peroxide ligand connection as has been
discussed in Sections 7.8.2 and 7.8.3. The reference 138 authors consider the
peroxide bridge unlikely because, for one reason, the metal – O and O – O bond
distances in the bovine heart enzyme are considerably longer than that found
for similar synthetic complexes (see reference 138 , Table 2; also see Table 7.9 ).
Typical bond distances for O 2 , O 2 − (superoxo), and O 22 − (peroxo) are 1.21,
1.28 – 1.30, and 1.40 – 1.50 Å , respectively, while the O – O bond distance reported
for PDB: 2OCC is equal to 1.63 Å.
Not discussed in the previous section were bovine heart crystal structures
featuring heme a 3 – Cu B with CO (carbon monoxide) - and N 3 − (azide) - bound
ligands, although these are common ligands in synthetic models for the oxi-
dized center. (See reference 137c .) Other CcO enzyme X - ray structures of P.
dentrifi cans (PDB: 1AR1),^156 R. sphaeroides (PDB: 1M56, 1M57),^144 and T.
thermophilus (PDB: 1EHK)^157 have proposed various connectors: μ - oxo and
μ - hydroxo species, water molecules, or some combination of these for the
heme a 3 – Cu B binuclear center. Because of their atomic resolution, the X - ray
structures are not completely clear as to whether the connector species do, in
fact, connect or are simply the nearby ligands of the two metal centers. (See
proposedR, A, P, F and O states as shown in Figures 7.40 and 7.42 .) The ref-