Tiny Peptide Model Shows How Cobalt and Cysteine Oxidize Together in Nitrile Hydratase Maturation

Nitrile hydratase is one of the most industrially valuable enzymes in biocatalysis, used at enormous scale to convert nitriles into amides such as acrylamide and nicotinamide under mild conditions. At the heart of each enzyme sits one of the strangest metal centers in biology: a low-spin iron(III) or cobalt(III) ion gripped by two backbone amide nitrogens and three cysteine-derived sulfur ligands, two of which are chemically modified after the protein is built into sulfenic and sulfinic acids. This unusual “claw-setting” coordination environment is essential for catalysis, yet how the enzyme assembles it — how the metal is inserted and how the cysteines become oxidized in the right sequence — has remained poorly understood at the atomic level. A new study published in Frontiers in Chemical Biology tackles that question with a deliberately minimal model: a synthetic eight-residue peptide carrying the conserved metal-binding sequence of the enzyme.

The research team, led by Irene R. A. M. Ogutu of Regis University and Richard C. Holz of the Colorado School of Mines, together with colleagues at Marquette University, synthesized a peptide with the sequence Ac-VCTLCSCY. The sequence derives from the cobalt-type nitrile hydratase of Pseudonocardia thermophila JCM 3095 and follows the conserved CXXCSCX motif found across nitrile hydratases. An acetyl group was attached to the N-terminus to improve solubility, and the peptide was prepared at greater than 98.5 percent purity. The strategy builds on earlier work showing that the same motif could be used to generate active-site biomimetic complexes, including an iron-bound version whose reaction with oxygen was previously characterized.

Under strictly anaerobic conditions in a glove box, the researchers added one equivalent of cobalt(II) chloride to the peptide dissolved in degassed buffer at pH 8.0. The solution turned light green, and ultraviolet-visible absorption spectroscopy revealed intense bands at 305 and 355 nanometers with molar absorptivities of roughly 2,800 and 2,180 per molar per centimeter. These bands are characteristic of ligand-to-metal charge transfer from cysteine thiolate ligands bound to cobalt(II). Because each thiolate–cobalt(II) interaction is known to contribute approximately 900 to 1,200 units of molar absorptivity to such bands, the observed intensities pointed to two cysteine thiolate ligands coordinated to the metal center within the peptide complex.

Titration experiments monitoring the absorbance near 685 nanometers showed that approximately one cobalt(II) ion binds per peptide, with an apparent dissociation constant of 0.45 micromolar. The authors are careful to note a limitation of this measurement: because the peptide concentration in the titration greatly exceeded the apparent dissociation constant, the experiment establishes tight, approximately stoichiometric binding but does not rigorously determine a sub-micromolar dissociation constant. The binding strength is nonetheless comparable to values reported for related biomimetic peptide complexes such as Co-NiSOD and Co-SODA, supporting the idea that a single divalent metal ion occupies the motif.

The most striking spectroscopic result came from electron paramagnetic resonance. X-band EPR spectra recorded at 12 kelvin showed two distinct signals from the anaerobic cobalt(II)-peptide complex. One, with g-values of 2.30, 2.25, and 2.03, indicated a low-spin cobalt(II) ion with its unpaired electron in a 3d z-squared orbital. The other, a low-amplitude resonance between 1,000 and 2,000 gauss with turning points near g-prime values of 5.7 and 4.8, indicated a high-spin cobalt(II) center. Double integration of the spectra showed that the high-spin and low-spin signals carried approximately equal relative integrated intensities. Measurements at L-band frequency, 1.3 gigahertz, resolved the hyperfine structure of the low-spin species more clearly, since the lower frequency collapses strain in the g-tensor. The coexistence of two spin states, the authors suggest, likely reflects the flexibility of the small peptide scaffold, which can accommodate multiple coordination geometries around the metal.

To pin down those geometries, the team turned to computational chemistry. They built six density functional theory models of the cobalt(II)-peptide, three for each spin state, differing in which of the three cysteine residues — Cys2, Cys5, or Cys7 — remained a neutral, weakly or non-coordinating thiol while the other two served as anionic thiolates. Geometry optimizations used the BP86 functional with a def2-SVP basis set, dispersion corrections, and an aqueous continuum model, with two explicit water molecules included to capture hydrogen bonding to the cysteine and serine side chains. Time-dependent DFT calculations reproduced the near-ultraviolet charge-transfer transitions with molar absorptivities of 2,000 to 5,000, in reasonable agreement with experiment. For the high-spin species, the computed ligand-field intensities of the four-coordinate model matched the experimental values, while five-coordinate models fell far short, leading the authors to favor a distorted tetrahedral arrangement with two amidate nitrogens and the thiolates of Cys2 and Cys5, leaving the Cys7 thiol uncoordinated.

The low-spin species required a different approach. The team calculated EPR g-values using a multiconfigurational method pairing complete active space self-consistent field calculations with N-electron valence perturbation theory, using a CAS(7,5) active space of seven electrons in the cobalt 3d shell. The four-coordinate square-planar model produced wildly anisotropic g-values of 5.41, 0.49, and 0.46, typical of synthetic four-coordinate low-spin cobalt(II) complexes and clearly inconsistent with the modestly anisotropic experimental signal. The five-coordinate models, by contrast, gave g-values between 1.98 and 2.57, close to the experimental range of roughly 2.03 to 2.40, with the model placing the neutral thiol at Cys7 reproducing the axial character of the g-tensor especially well. The favored picture for the low-spin species is therefore a five-coordinate, distorted square-pyramidal geometry, with the unpaired electron in a 3d z-squared orbital perpendicular to the equatorial cobalt-nitrogen-sulfur plane.

Then came the oxygen. When the anaerobic green solution was exposed to air, it darkened to green-brown and, over about five hours, turned orange-brown. The ligand-field bands between 600 and 750 nanometers faded and vanished, replaced by new absorption bands near 350, 465, and 620 nanometers, characteristic of thiolate-to-cobalt(III) charge transfer. In parallel, the EPR signals of both cobalt(II) species were rapidly extinguished, consistent with oxidation to a low-spin cobalt(III) center that is diamagnetic and therefore EPR-silent — exactly the configuration found in the mature cobalt-type enzyme. Control experiments showed that the high-spin EPR signal of free cobalt(II) in buffer is distinct from that of the peptide complex and that air exposure did not simply liberate cobalt from the peptide, ruling out a trivial dissociation explanation.

Mass spectrometry added the ligand side of the story. Quantitative time-of-flight liquid chromatography–mass spectrometry of samples exposed to air for 3, 10, and 15 hours revealed a prominent ion at mass-to-charge ratio near 1,074 after three hours, corresponding to the cobalt-bound peptide carrying three additional oxygen atoms relative to the reduced complex, alongside a potassium adduct commonly seen in such measurements. The unoxidized cobalt-peptide peak near 986 remained visible but diminished in relative intensity as oxidation time increased. The authors interpret the mass shifts as consistent with oxidation of cysteine residues to sulfenic and sulfinic acid derivatives, the same post-translational modifications found in the genuine enzyme active site, while cautioning that mass spectrometry alone cannot establish the metal’s oxidation state and that the spectra are complex.

The broader significance lies in what the minimal model proves about the maturation process. In the natural enzyme, a small activator protein acts as a metallochaperone for the alpha subunit, and a proposed alpha-subunit swapping mechanism for P. thermophila nitrile hydratase holds that cobalt(II) is inserted first and then oxidized, along with the equatorial cysteines, to yield the functional low-spin cobalt(III) center bearing sulfenic and sulfinic acid ligands. What the new study demonstrates is that the conserved metal-binding motif alone — without the rest of the protein — already contains sufficient chemical information to bind cobalt(II) in two distinct precursor geometries and to undergo coupled metal and cysteine oxidation upon reaction with dioxygen. The authors note that the order of metal and ligand oxidation cannot be established from the present data, only that both occur during the reaction, and they suggest that prolonged oxygen exposure might generate more extensively oxidized species, potentially including a bis-sulfinic-acid environment analogous to overoxidized, inactive enzyme. The cobalt chemistry closely parallels what was previously found for the iron-bound version of the same peptide, hinting at a shared maturation logic across the two nitrile hydratase types. As a caveat, this is a model system of a single short peptide in solution, not the enzyme itself, so it captures the intrinsic reactivity of the motif rather than the full, activator-protein-guided process in living cells. Even so, the finding that a mere eight residues can support both precursor states and their oxygen-driven conversion brings chemists a step closer to understanding how nature builds one of its most unusual catalytic metal centers.

Subject of Research: Biomimetic metallopeptide studies of cobalt nitrile hydratase active-site maturation

Article Title: A metallopeptide model for the active site of cobalt nitrile hydratase reveals coupled metal and cysteine oxidation during active-site maturation

Article References: Ogutu, I. R. A. M., Miller, C. G., Hebbeler, N., Luong, A., Fiedler, A. T., Bennett, B., & Holz, R. C. (2026). A metallopeptide model for the active site of cobalt nitrile hydratase reveals coupled metal and cysteine oxidation during active-site maturation. Frontiers in Chemical Biology, 5, Article 1980777. https://doi.org/10.3389/fchbi.2026.1980777

Image Credits: AI Generated

DOI: 10.3389/fchbi.2026.1980777

Keywords: nitrile hydratase, cobalt, metalloenzyme, active-site maturation, metallopeptide, EPR spectroscopy, DFT calculations, cysteine oxidation, bioinorganic chemistry, mass spectrometry, Co(III), biocatalysis