Fluorescence labeling methods influence the aggregation process of α-syn in vitro differently.

In a previous study, the coexistence of different aggregation pathways of insulin and ß-amyloid (Aß) peptides was demonstrated by correlative stimulated emission depletion (STED) microscopy and atomic force microscopy (AFM). This had been explained by suboptimal proteins labeling strategies that generate heterogeneous populations of aggregating species. However, because of the limited number of proteins considered, the failure of the fluorescent labeling that occurs in a large portion of the aggregating fibrils observed for insulin and Aß peptides, could not be considered a general phenomenon valid for all molecular systems. Here, we investigated the aggregation process of α-synuclein (α-syn), an amyloidogenic peptide involved in Parkinson's disease, which is significantly larger (MW ∼14 kDa) than insulin and Aß, previously investigated. The results showed that an unspecific labeling procedure, such as that previously adopted for shorter proteins, reproduced the coexistence of labeled/unlabeled fibers. Therefore, a site-specific labeling method was developed to target a domain of the peptide scarcely involved in the aggregation process. Correlative STED-AFM illustrated that all fibrillar aggregates derived from the aggregation of α-syn at the dye-to-protein ratio of 1 : 22 were fluorescent. These results, demonstrated here for the specific case of α-syn, highlight that the labeling artifacts can be avoided by careful designing the labeling strategy for the molecular system under investigation. The use of a label-free correlative microscopy technique would play a crucial role in the control of the setting of these conditions.

DOPAL derived alpha-synuclein oligomers impair synaptic vesicles physiological function.

Parkinson's disease is a neurodegenerative disorder characterized by the death of dopaminergic neurons and by accumulation of alpha-synuclein (aS) aggregates in the surviving neurons. The dopamine catabolite 3,4-dihydroxyphenylacetaldehyde (DOPAL) is a highly reactive and toxic molecule that leads to aS oligomerization by covalent modifications to lysine residues. Here we show that DOPAL-induced aS oligomer formation in neurons is associated with damage of synaptic vesicles, and with alterations in the synaptic vesicles pools. To investigate the molecular mechanism that leads to synaptic impairment, we first aimed to characterize the biochemical and biophysical properties of the aS-DOPAL oligomers; heterogeneous ensembles of macromolecules able to permeabilise cholesterol-containing lipid membranes. aS-DOPAL oligomers can induce dopamine leak in an in vitro model of synaptic vesicles and in cellular models. The dopamine released, after conversion to DOPAL in the cytoplasm, could trigger a noxious cycle that further fuels the formation of aS-DOPAL oligomers, inducing neurodegeneration.

Binding Interactions of Agents That Alter α-Synuclein Aggregation.

Further examination of peptides with well-folded antiparallel ß strands as inhibitors of amyloid formation from α-synuclein has resulted in more potent inhibitors. Several of these had multiple Tyr residues and represent a new lead for inhibitor design by small peptides that do not divert α-synuclein to non-amyloid aggregate formation. The most potent inhibitor obtained in this study is a backbone cyclized version of a previously studied ß hairpin, designated as WW2, with a cross-strand Trp/Trp cluster. The cyclization was accomplished by adding a d-Pro-l-Pro turn locus across strand termini. At a 2:1 peptide to α-synuclein ratio, cyclo-WW2 displays complete inhibition of ß-structure formation. Trp-bearing antiparallel ß-sheets held together by a disulphide bond are also potent inhibitors. 15N HSQC spectra of α-synuclein provided new mechanistic details. The time course of 15N HSQC spectral changes observed during ß-oligomer formation has revealed which segments of the structure become part of the rigid core of an oligomer at early stages of amyloidogenesis and that the C-terminus remains fully flexible throughout the process. All of the effective peptide inhibitors display binding-associated titration shifts in 15N HSQC spectra of α-synuclein in the C-terminal Q109-E137 segment. Cyclo-WW2, the most potent inhibitor, also displays titration shifts in the G41-T54 span of α-synuclein, an additional binding site. The earliest aggregation event appears to be centered about H50 which is also a binding site for our most potent inhibitor.

Number and Brightness analysis of alpha-synuclein oligomerization and the associated mitochondrial morphology alterations in live cells.

BACKGROUND: Alpha-synuclein oligomerization is associated to Parkinson's disease etiopathogenesis. The study of alpha-synuclein oligomerization properties in live cell and the definition of their effects on cellular viability are among fields expected to provide the knowledge required to unravel the mechanism(s) of toxicity that lead to the disease. METHODS: We used Number and Brightness method, which is a method based on fluorescence fluctuation analysis, to monitor alpha-synuclein tagged with EGFP aggregation in living SH-SY5Y cells. The presence of alpha-synuclein oligomers detected with this method was associated with intracellular structure conditions, evaluated by fluorescence confocal imaging. RESULTS: Cells overexpressing alpha-synuclein-EGFP present a heterogeneous ensemble of oligomers constituted by less than 10 monomers, when the protein approaches a threshold concentration value of about 90nM in the cell cytoplasm. We show that the oligomeric species are partially sequestered by lysosomes and that the mitochondria morphology is altered in cells presenting oligomers, suggesting that these mitochondria may be dysfunctional. CONCLUSIONS: We showed that alpha-synuclein overexpression in SH-SY5Y causes the formation of alpha-synuclein oligomeric species, whose presence is associated with mitochondrial fragmentation and autophagic-lysosomal pathway activation in live cells. GENERAL SIGNIFICANCE: The unique capability provided by the Number and Brightness analysis to study alpha-synuclein oligomer distribution and properties, and the study of their association to intracellular components in single live cells is important to forward our understanding of the molecular mechanisms of Parkinson's disease and it may be of general significance when applied to the study of other aggregating proteins in cellular models.

Role of the tertiary structure in the diphenol oxidase activity of Octopus vulgaris hemocyanin.

The functional differences between the oxygen transport protein Hemocyanin and the enzymes Tyrosinase and Catechol oxidase are believed to be governed, at least in part, by the tertiary structure, which differs in these molecules and controls the accessibility of their copper containing active site for substrate(s). Accordingly, Octopus vulgaris Hemocyanin catalyses the o-diphenol oxidation to o-quinone at a very low rate. The crystallographic structure of one of the functional units (called Odg) of O. dofleini Hemocyanin shows two domains, a mainly alpha-helical domain that directly binds the copper ions of the reaction center and a beta-strand domain that precludes access to the active site to ligands bigger than molecular oxygen. In this work, we have first cleaved the whole protein and then purified different oxygen binding functional units from O. vulgaris Hemocyanin. These functional units were used in activity assays with l-DOPA, the paradigmatic substrate for Catechol oxidase. All functional units show a negligible enzymatic activity. The procedure to generate the functional units induces in only one of them a proteolytic cleavage. Amino terminal sequencing and mass spectroscopy of the fragments allow to place the cleavage site between the alpha and beta domains of the functional unit homologous to Odd, in the O. dofleini sequence. An increase, up to three orders of magnitude, of Tyrosinase-like activity was observed when the cleaved Odd-like was incubated with the substrate in the presence of trifluoroethanol or hexafluoroisopropanol.

Structural role of the copper ions in the dinuclear active site of Carcinus aestuarii hemocyanin.

In this work we show, by a combination of biochemical and biophysical approaches, that the copper ions bound in the binuclear active site of Carcinus aestuarii hemocyanin play a stabilizing role on the tertiary structure of the protein. Upon removal of copper, the monomeric hemocyanin, but not the hexameric oligomer, undergoes changes at the level of tertiary structure while the secondary structure is almost unaffected. By Small-Angle X-Ray Scattering, supported by gel chromatography measurements, it can be concluded that the apo-monomer, but not the holo form or the hexameric form, undergoes a slow time-dependent oligomerization process.

Structural properties, conformational stability and oxygen binding properties of Penaeus monodon hemocyanin.

Hemocyanin sequences allineament shows the presence of highly invariant regions especially in the active site and in the tight intersubunits interaction sites. Comparing the aminoacids in contact regions between monomers is possible to interpret the stability of hexamers.

Comparison of the X-ray absorption properties of the binuclear active site of molluscan and arthropodan hemocyanins.

The structural characteristics of oxy- and deoxy-hemocyanins have been investigated using X-ray absorption spectroscopy both in the near-edge (XANES) and for the first shell contribution in the EXAFS region. Several arthropodan and molluscan hemocyanins have been studied in order to trace the inter- and intra-phyla differences. The XANES spectra of oxy-hemocyanins of the different species are remarkably similar, consistent with a very strongly conserved co-ordination geometry of the copper active site. In contrast, small but significant differences are observed between the deoxy-forms of arthropodan and molluscan proteins. In particular, the XANES spectra of deoxy-arthropodan hemocyanins (with the exception of L. polyphemus Hc) show a more intense edge feature at approximately 8983 eV. This difference is tentatively assigned to a more planar geometry of the copper-ligands system in the arthropodan rather than in the molluscan proteins. The first shell analysis of the EXAFS modulation is consistent with the presence of n=3Nepsilon(2) imidazole nitrogens at an average distance of 1.92 +/- 0.03 A from copper in all the deoxy-hemocyanins investigated. Binding of dioxygen results for all hemocyanins in the increase of the number of first shell back-scattering atoms to n=5 with average distances of 1.93 A. Alternatively, by separating the contribution of Nepsilon(2) imidazole nitrogens and of peroxide O-atoms, n=3 ligands at 1.98 +/- 0.03 A and n=2 ligands at 1.87 +/- 0.03 A are found.

Molecular heterogeneity of the hemocyanin isolated from the king crab Paralithodes camtschaticae.

Native Paralithodes camtschaticae hemocyanin is found as a mixture of dodecamers (24S; 80%) and hexamers (16S; 20%). Removal of Ca2+ ions by dialysis against EDTA-containing buffer solution at neutral pH induces complete dissociation of the 24S form into the 16S form. Under these conditions, a further increase in pH to 9.2 produces complete dissociation of the hexamers into monomers (5S). In both cases, the dissociation process is reversible. The dodecamer (24S) is composed of two different hexamers which can be discriminated only by ion-exchange chromatography in the presence of Ca2+ ions. At alkaline pH and in the presence of EDTA, two major monomeric fractions can be separated by ion-exchange chromatography: ParcI (60%) and ParcII (40%). The reassociation properties of the two fractions were studied separately to define their ability to form hexamers and dodecamers. The oxygen-binding properties of the different aggregation states were investigated. Native hemocyanin binds O2 co-operatively (nH = 3) and with low affinity (p50 approximately 103 Torr). The two monomeric fractions, ParcI and ParcII, are not co-operative and the affinity is twice that of the native protein (p50 approximately 65 and 52 Torr). Oxygen-binding measurements of native hemocyanin carried out at different pH values indicate a strong positive Bohr effect within the pH range 6.5-8.0 and an increase in oxygen affinity at pH below 6.5.

Saccharose solid matrix embedded proteins: a new method for sample preparation for X-ray absorption spectroscopy.

In this study, solid samples of hemoglobin and hemocyanin have been prepared by embedding the proteins into a saccharose-based matrix. These materials have been developed specifically for specimens for X-ray absorption spectroscopy (XAS). The preservation of protein conformation and active site organization was tested, making comparisons between the solid and the corresponding liquid samples, using resonance Raman, infra red, fluorescence and XAS. The XAS spectra of irradiated solid and liquid samples were then compared, and the preservation of biological activity of the proteins during both preparation procedure and X-ray irradiation was assessed. In all cases, the measurements clearly demonstrate that protein solid samples are both structurally and functionally quite well preserved, much better than those in the liquid state. The saccharose matrix provides an excellent protection against X-ray damages, allowing for longer exposure to the X-ray beam. Moreover, the demonstrated long-term stability of samples permits their preparation and storage in optimal conditions, allowing for the repetition of data collection with the same sample in several experimental sessions. The very high protein concentration that can be reached results in a significantly better signal-to-noise ratio, particularly useful for high molecular weight proteins with a low metal-to-protein ratio. On the bases of the above-mentioned results, we propose the new method as a standard procedure for the preparation of biological samples to be used for XAS spectroscopy.

EPR study of the dinuclear active copper site of tyrosinase from Streptomyces antibioticus.

The [Cu(I)-Cu(II)] half-met form of the dinuclear copper site of tyrosinase has been probed by continuous wave electron paramagnetic resonance (EPR) and hyperfine sublevel correlation (HYSCORE) spectroscopy in the presence and absence of inhibitors. In all cases the EPR spectrum is indicative of a d(x(2)-y(2)) ground state for the unpaired electron. From the cross-peaks observed in the HYSCORE spectra, proton hyperfine coupling constants were obtained that are compatible with a hydroxide ion in an equatorial coordination position of the paramagnetic copper. After changing the water solvent to D(2)O or after addition of the inhibitors p-nitrophenol or L-mimosine, the proton signals disappear. The relevance of these findings for understanding the catalytic cycle is discussed.

Interaction and coordination geometries for Ag(I) in the two metal sites of hemocyanin.

111Ag(I) perturbed angular correlations of gamma-rays (PAC) has been used to investigate the binuclear metal site of 111Ag(I)-substituted Carcinus aestuarii deoxyhemocyanin. The studies have shown that apo-hemocyanin is able to bind 2 mol of Ag(I) per mol of protein and that the binding is specific for the metal ion sites. The PAC spectra show pronounced changes when the stoichiometry of Ag(I) to protein is increased from 0.1 to 2.0. These changes have been interpreted as evidence of interactions between the two sites in terms of a structural destabilization of the first occupied site caused by the occupation of the second site. The experimental data for the Ag(I)-substituted metal sites do not agree well with the three-coordinated structure found in the Cu(I) holo-protein. However, if a water molecule is included as a coordinating ligand in the Ag(I) metal site a successful interpretation of the experimental data can be obtained.

1H NMR spectroscopy of the binuclear Cu(II) active site of Streptomyces antibioticus tyrosinase.

The 600 MHz 1H NMR spectrum of tyrosinase (31 kDa) of Streptomyces antibioticus in the oxidized, chloride-bound form is reported. The downfield part of the spectrum (15-55 ppm) exhibits a large number of paramagnetically shifted signals. The paramagnetism is ascribed to a thermally populated triplet state. The signals derive from six histidines binding to the metals through their Nepsilon atoms. There is no evidence for endogenous bridges. The exchange coupling, -2J, amounts to 298 cm(-1). In the absence of chloride the peaks broaden. This is ascribed to a slowing down of the electronic relaxation. The exchange coupling decreases to -2J=103 cm(-1).

Understanding the electronic properties of the CuA site from the soluble domain of cytochrome c oxidase through paramagnetic 1H NMR.

The soluble domain of the subunit II of cytochrome c oxidase from Paracoccus versutus was cloned, expressed, and studied by 1H NMR at 600 MHz. The properties of the redox-active dinuclear CuA site in the paramagnetic mixed-valence Cu(I)-Cu(II) state were investigated in detail. A group of relatively sharp signals found between 30 and 15 ppm in the 1H NMR spectrum correspond to the imidazole protons of the coordinated histidines (H181 and H224). A second group of broader and farther shifted signals between 50 and 300 ppm are assigned to Hbeta protons of the bridging cysteines (C216 and C220); the protons from the weak M227 and E218 ligands do not shift outside of the diamagnetic envelope. About 40% of the total spin density appears delocalized over the cysteine-bridging ligands while a much smaller amount is delocalized on the two ligand histidines. The latter have similar spin density distributions. Analysis of the pattern of the hyperfine shifts of the Cys H beta protons shows that the ground state bears 2B3u character, in which the sulfur lobes in the singly occupied molecular orbital are aligned with the Cu-Cu axis. Analysis of the temperature dependence of the shifts of the Cys H beta signals leads to the conclusion that the 2B2u excited state is thermally accessible at room temperature (Delta E approximately kT).

The binding of azide to copper-containing and cobalt-containing forms of hemocyanin from the mediterranean crab Carcinus aestuarii.

To establish the competence of the active site of hemocyanin to acquire diverse coordination geometries, the binding of azide to three forms of a crab hemocyanin, the dinuclear cupric or met-hemocyanin, the mononuclear cupric or met-apo-hemocyanin, and the mononuclear Co(II)-substituted derivative has been studied by near-ultraviolet circular dichroism and EPR spectroscopies. The near-ultraviolet circular dichroism spectra of the various derivatives present qualitatively similar features, namely a negative peak around 335 nm in the case of the two copper-containing derivatives and a three-component pattern with the Co(II) derivative. Upon decreasing the pH from 7.0 to 5.5 a decrease of optical activity is observed with all protein samples. The characteristic CD features, attributable to N(imidazole)-to-metal and to OH -to-metal charge-transfer transitions, are strongly affected by azide binding. In particular, the intensity of the negative band exhibited by the two copper-containing protein forms decreases with the onset of a new negative feature with maximum around 400 nm diagnostic for azide-to-Cu(II) charge-transfer transitions. The visible region is affected as well, indicating that changes in the coordination sphere of copper take place. The affinity for azide of the different protein forms is higher at low pH. EPR measurements on the paramagnetic met-apo-hemocyanin derivative as a function of pH demonstrate heterogeneity in the coordination environment at low pH. In the presence of azide an increase of rhombic distortion of the EPR spectra is observed and on the basis of the identified sets of copper hyperfine features in the course of azide titration experiments two different azide bound forms of met-apo-hemocyanin can be detected. The CD and EPR data at the different pH values are consistent with a reaction scheme in which azide replaces a fourth ligand in the metal-coordination sphere, identified as a water or hydroxide molecule.

The oxidation of hemocyanin. Kinetics, reaction mechanism and characterization of met-hemocyanin product.

The reaction that gives met-hemocyanin from Octopus vulgaris oxy-hemocyanin has been reinvestigated under several experimental conditions. Various anions including azide, fluoride and acetate have been found to promote this reaction. Kinetic data indicate that the reaction mechanism is different from that currently accepted involving a peroxide displacement of bound dioxygen through an associative chemistry on an open axial position of the copper ions [Hepp, A. F., Himmelwright, R. S., Eickman, N. C. & Solomon, E. I. (1979) Biochem. Biophys. Res. Commun. 89, 1050-1057; Solomon, E. I. in Copper proteins (Spiro, T. G., ed.) pp. 43-108, J. Wiley, New York]. Our study suggests that the protonated form of the anion is likely to be the species reacting with the oxygenated form of the protein. Furthermore, it is also proposed that protonation of bound dioxygen generates an intermediate hydroperoxo-dicopper(II) complex to which the exogenous anion is also bound. This intermediate in not accumulated and preceds the release of hydrogen peroxide by reaction with water. Upon dialysis it leads to the met-hemocyanin form. The structure of this dinuclear copper(II) derivative contains a di-mu-hydroxo bridge but there is evidence from optical and circular dichroism spectra for partial protonation of these bridges at low pH. As a consequence, while one azide molecule binds in the bridging mode to met-hemocyanin with low affinity (K = 30 M-1) at pH 7.0, it binds with much higher affinity at pH 5.5 (K = 1500 M-1), where a second azide ligand also binds in the terminal mode (K = 20 M-1). The coordination mode of the azide ligands is deduced from the optical and circular dichroism spectra of the protein complexes.

Cu(II) coordination in arthropod and mollusk green half-methemocyanins analyzed by electron spin-echo envelope modulation spectroscopy.

Hemocyanin (Hc) is a dinuclear copper protein that binds oxygen reversibly. The structure of the Cu(II) site in a derivative of hemocyanin known as green half-met (GHM) has been analyzed using the pulsed EPR technique of electron spin-echo envelope modulation (ESEEM) spectroscopy. The derivative, prepared by treating the native protein with nitrite at low pH, contains a mixed-valent binuclear copper center. It was shown through chemical assays and the ligand exchange reaction products identified by EPR spectroscopy to contain a nitrite ligand bound to Cu(II). The ESEEM spectra of green half-methemocyanins from mollusks and arthropods indicated that three imidazole ligands are coordinated to Cu(II). Therefore, a tetragonal N3O ligand structure (O is an oxygen of nitrite) is proposed. For GHM Hc from the mollusks Octopus vulgaris and Rapana thomasiana, the isotropic nitrogen nuclear hyperfine coupling constant, aiso, for the N delta (or remote) nitrogen of two imidazoles was approximately 1.4 MHz, while for the third, aiso congruent to 2.2 MHz. The difference between the two weaker nitrogens and the single, more strongly coupled nitrogen was smaller by 0.2 MHz in the GHM Hcs from the arthropods Carcinus maenas, Homarus americanus and Panulirus interruptus. The nitrogen nuclear quadrupole coupling constants and asymmetry parameters, e2Qq and eta, for the N delta nitrogens in nearly all cases were near 1.4 MHz and 0.8, respectively, although Rapana thomasiana GHM Hc exhibited a reduction in eta that may indicate weaker hydrogen bonding in the active site of this protein. The g and ACu (copper nuclear hyperfine coupling) values for the derivatives, and the finding of three similar nuclear hyperfine coupling constants for the N delta sites of imidazole ligands, when considered with the orientation-specific information obtained using angle-selection methods for simulation of ESEEM spectra, suggest a distorted tetragonal Cu(II) structure in which three imidazoles and a nitrite ligand are bound near the equatorial plane. The finding that the two molluscan GHM Hcs exhibit differences associated with the remote nitrogen of imidazoles bound to Cu(II) may be related to a structural variability in the active sites of these proteins not found in the arthropodan GHM Hcs examined.

Structural characterization of mononuclear Cu(II) and its nitrite complex in the active site of Carcinus maenas hemocyanin.

The preparation of a mononuclear Cu(II) derivative of Carcinus maenas hemocyanin (Cu(II)-Hc) and a nitrite complex of the derivative (Cu(II)-Hc-NO2-) are described. Several techniques have been used in their characterization, including X-ray absorption, continuous wave (cw) EPR, and electron spin-echo envelope modulation (ESEEM) spectroscopies. EXAFS results for Cu(II)-Hc indicate the presence of three ligands at 1.99 +/- 0.01 A and a fourth one at 2.26 +/- 0.01 A from the copper. The same coordination number and very similar bond lengths were obtained for Cu(II)-Hc-NO2-. On the basis of simulations of three-pulse ESEEM spectra, three equivalent imidazole nitrogens coupled to Cu(II) were identified in Cu(II)-Hc. Upon the binding of nitrite, a decrease in the hyperfine interaction for two of the three imidazole nitrogens was observed by ESSEM. Further, the results of a two-pulse ESEEM experiment are consistent with the assignment of the protons of a water ligand to Cu(II), which is displaced when nitrite is added. An analysis of X-ray absorption K-edge spectra suggests a coordination geometry intermediate between square-planar and tetrahedral for the metal centers in Cu(II)-Hc and Cu(II)-Hc-NO2-, in agreement with the g and ACu values determined by cw-EPR. On the basis of these results, an equivalent structure is suggested for Cu(II)-Hc-NO2- and the Cu(II) site in green half-methemocyanin, a partially oxidized binuclear derivative formed in the reaction of the native protein with nitrite.

The binding of Cd(II) to the hemocyanin of the Mediterranean crab Carcinus maenas.

The interaction of Carcinus hemocyanin with Cd(II) was studied. The incubation of the apoprotein with the metal yields a derivative containing 1 g-at. of EDTA-stable Cd(II) per 75 kDa. Spectroscopic data ruled out Cd(II) coordination to tryptophan or cysteine residues. The optical activity and fluorescence properties of the protein are affected by Cd(II) binding and indicate a rearrangement of tryptophan residues. The poor Cd(II) binding to the oxy-form and the resistance of Cd(II)-hemocyanin to EDTA treatment and to the regeneration by Cu(I) strongly indicate that Cd(II) binding to apohemocyanin occurs at the copper-free active site. During the metal-binding process, a marked increase of light scattering is observed. This effect, however, is reversible provided that the incubation medium contains SCN- and glycine as exogenous ligands of the metal in the bulk solution.

Preparation and spectroscopic characterization of a coupled binuclear center in cobalt(II)-substituted hemocyanin.

A binuclear cobalt derivative of arthropod hemocyanin (Hc) has been prepared by the reaction of apo-Hc with Co(II) in the presence of thiocyanate. The crude product of the reaction contains specifically and adventitiously bound metal, the latter being removable by EDTA treatment. The specifically bound Co(II) constitutes a binuclear metal center that exhibits optical and CD spectra typical in their absorption maxima and extinction coefficients of Co(II) complexes with near-tetrahedral geometry. The EPR spectrum of the binuclear Co(II) derivative contains a resonance at g approximately 13, which is characteristic of integer spin systems and indicates coupled metal ions; the excess Co(II) bound to crude products exhibits an EPR signal at g approximately 4. The time course of derivative formation was followed by EPR, optical and atomic absorption techniques, and by fluorimetry. The intensity of the optical absorption in the visible region due to Co(II) increases with increasing stoichiometry of specifically bound metal [up to 2 Co(II) per protein monomer], but the intensity of the Co(II) EPR signal increases only during the formation of a mononuclear derivative. As the reaction proceeds over approximately 100 h to the formation of the binuclear derivative, the EPR signal intensity decreases to 10% of the value expected for 2 mol of EPR-active Co(II)/mol of protein. The binuclear cobalt derivative cannot be reconstituted to native Hc with Cu(I), indicating the stable loading of Co(II) in the active site. EPR and optical spectroscopic evidence is presented showing that the binuclear derivative does not bind oxygen.