br Abbreviations br Acknowledgements br Introduction
Introduction Dopamine β-hydroxylase (DbH) is the enzyme catalysing the conversion of dopamine to norepinephrine, two important neurotransmitters involved in the central nervous system . This enzyme has been isolated from bovine chromaffin granules as a tetramer of 290 kDa , and exists as both membrane-associated and soluble forms. It is a copper protein  that contains two Type II copper atoms per active site, which both are Cu(II) in the resting state. No crystallographic data are available to date concerning DbH. However some indications about its structure have been given by EXAFS measurements , and by the examination of the recent publication of peptidyl glycine α-amidating monooxygenase hydroxylating domain (PHM) structure . PHM, which was shown to possess mechanistic similarities with DbH ,  and a sequence resembling that of DbH , contains two copper atoms 11 Å distant from each other, separated by a solvent pocket. In DbH, one copper (CuA) is bound to three histidine and a water molecule, while the other (CuB) is bound to two histidine residues and one additional methionine residue in the reduced form . No magnetic coupling exists between the two copper ions, suggesting a distance more than 6 Å . This would exclude the possibility of a transient Cu2-bis(μ-oxo) active species, which is observed in various model compounds . During catalysis, DbH copper atoms cycle between Cu(II) and Cu(I), as shown by EPR measurements . A commonly used reducing cofactor is ascorbate, also present in chromaffin granules, and suggested to be a cofactor of DbH in vivo. Oxygen is the second cofactor for DbH, which accepts dopamine, tyramine or 2-phenyl-ethylamine as substrates. As early as 1961, Levin and Kaufman showed that incubation of the enzyme with either reducer ascorbate in the presence of air or oxidant H2O2 was responsible for a loss of catalytic activity . Kinetic measurements on the influence of inactivating species concentration onto activity have later been realised by Skotland and Ljones . Up to now, no mechanistic study was able to provide an explanation for such losses of activity. Reduction of the two copper atoms results in the activation of dioxygen by DbH and the generation of a peroxo-copper intermediate at CuB, which undergoes a cleavage of the oxygen–oxygen bond . As a next step, a hydrogen p-selectin is abstracted from the substrate, before incorporation of the oxygen atom from the oxo-copper radical. The species responsible for this abstraction is not identified to date. Moreover it is not clear yet whether catalysis requires two copper atoms  or only one . A tyrosinyl radical was proposed by Klinman and co-workers to form transiently during catalysis  (Scheme 1A). Therefore we decided to investigate the degradation of a tyrosine residue following inactivation by ascorbate/oxygen or H2O2. In this article, we characterise the inactivated form of the enzyme produced by reaction with H2O2 or ascorbate/oxygen. We show that inactivation of DbH by hydrogen peroxide is partially prevented by substrate tyramine. Using visible spectroscopy and the quinone staining method by Paz et al. , we prove that inactivated DbH contains a protein-bound quinone derivative.
Materials and methods
Discussion NBT redox cycling staining was performed on ascorbate or hydrogen peroxide-inactivated DbH. This method proved that a protein-bound quinone derivative is present in inactivated enzyme. Our study and that describing this method  do not reveal any quinonoid component bound to active DbH. Therefore, the quinone derivative is generated during inactivation of the enzyme. Moreover, the absence of any NBT reduction by Apo-DbH demonstrates that copper is essential to the formation of this derivative. The presence of an oxidised phenolic side-chain is a proof for an oxidising attack on a tyrosine residue during DbH inactivation, thus suggesting the formation of a tyrosinyl radical during catalysis. Tyr residues have been shown to intervene as radical intermediates in other metalloenzymes, such as RNR . The nature of this oxidised species deserves discussion. Indeed, oxidation of a tyrosine residue can either generate a dopa-quinone (DPQ) or a TPQ, after insertion of one or two oxygen atoms on its side-chain. A protein-bound quinone has already been observed in other copper containing proteins. In copper monoamine oxidase, a carbonyl cofactor was evidenced, which was first proposed as a pyrroquinolinequinone  and later identified as a TPQ . It was later established that this essential cofactor was generated through a copper involving redox process . The presence of this modified residue can be let out both by visible absorption spectroscopy and using the NBT redox cycling test . In that paper, it is proved that the intensity of the blue colouration depends on the reducing quinone compound tested. For instance, DPQ has to be about 10 times as concentrated as TPQ to reduce identical amounts of NBT to formazan. In our redox cycling assays, we always obtained less intense formazan stains for DbH than for Cu-AO, despite using comparable amounts of proteins (220 vs 140 pmol of monomers). Such results suggest that the protein-bound quinone in inactivated DbH is not a TPQ but a DPQ. On the other hand, direct spectroscopic detection of the oxidised tyrosine residue in inactivated DbH discloses an absorption band centred at 408 nm, as opposed to the broad 474 nm band from MAO TPQ . Our value is rather close to that observed for free quinones, such as that produced by the oxidation of di-tert-butyl-catechol . On grounds of both redox cycling and visible spectroscopy results, we therefore propose that inactivation of DbH by ascorbate/oxygen or hydrogen peroxide generates a DPQ species that would be located nearby the active site copper responsible for dioxygen binding.