The role of the active site Cu(2+) of phenylethylamine oxidase from Arthrobacter globiformis (AGAO) has been studied by substitution with other divalent cations, where we were able to remove >99.5% of Cu(2+) from the active site. The enzymes reconstituted with Co(2+) and Ni(2+) (Co- and Ni-AGAO) exhibited 2.2 and 0.9% activities, respectively, of the original Cu(2+)-enzyme (Cu-AGAO), but their K(m) values for amine substrate and dioxygen were comparable. X-ray crystal structures of the Co- and Ni-AGAO were solved at 2.0-1.8 A resolution. These structures revealed changes in the metal coordination environment when compared to that of Cu-AGAO. However, the hydrogen-bonding network around the active site involving metal-coordinating and noncoordinating water molecules was preserved. Upon anaerobic mixing of the Cu-, Co-, and Ni-AGAO with amine substrate, the 480 nm absorption band characteristic of the oxidized form of the topaquinone cofactor (TPQ(ox)) disappeared rapidly (< 6 ms), yielding the aminoresorcinol form of the reduced cofactor (TPQ(amr)). In contrast to the substrate-reduced Cu-AGAO, the semiquinone radical (TPQ(sq)) was not detected in Co- and Ni-AGAO. Further, in the latter, TPQ(amr) reacted reversibly with the product aldehyde to form a species with a lambda(max) at around 350 nm that was assigned as the neutral form of the product Schiff base (TPQ(pim)). Introduction of dioxygen to the substrate-reduced Co- and Ni-AGAO resulted in the formation of a TPQ-related intermediate absorbing at around 360 nm, which was assigned to the neutral iminoquinone form of the 2e(-)-oxidized cofactor (TPQ(imq)) and which decayed concomitantly with the generation of TPQ(ox). The rate of TPQ(imq) formation and its subsequent decay in Co- and Ni-AGAO was slow when compared to those of the corresponding reactions in Cu-AGAO. The low catalytic activities of the metal-substituted enzymes are due to the impaired efficiencies of the oxidative half-reaction in the catalytic cycle of amine oxidation. On the basis of these results, we propose that the native Cu(2+) ion has essential roles such as catalyzing the electron transfer between TPQ(amr) and dioxygen, in part by providing a binding site for 1e(-)- and 2e(-)-reduced dioxygen species to be efficiently protonated and released and also preventing the back reaction between the product aldehyde and TPQ(amr).

The quinone cofactor TPQ in copper amine oxidase is generated by posttranslational modification of an active site tyrosine residue. Using X-ray crystallography, we have probed the copper-dependent autooxidation process of TPQ in the enzyme from Arthrobacter globiformis. Apo enzyme crystals were anaerobically soaked with copper; the structure determined from this crystal provides a view of the initial state: the unmodified tyrosine coordinated to the bound copper. Exposure of the copper-bound crystals to oxygen led to the formation of freeze-trapped intermediates; structural analyses indicate that these intermediates contain dihydroxyphenylalanine quinone and trihydroxyphenylalanine. These are the first visualized intermediates during TPQ biogenesis in copper amine oxidase.

Copper amine oxidase contains a post-translationally generated quinone cofactor, topa quinone (TPQ), which mediates electron transfer from the amine substrate to molecular oxygen. The overall catalytic reaction is divided into the former reductive and the latter oxidative half-reactions based on the redox state of TPQ. In the reductive half-reaction, substrate amine reacts with the C5 carbonyl group of the oxidized TPQ, forming the substrate Schiff base (TPQ(ssb)), which is then converted to the product Schiff base (TPQ(psb)). During this step, an invariant Asp residue with an elevated pKa is presumed to serve as a general base accepting the alpha proton of the substrate. When Asp298, the putative active-site base in the recombinant enzyme from Arthrobacter globiformis, was mutated into Ala, the catalytic efficiency dropped to a level of about 10(6) orders of magnitude smaller than the wild-type (WT) enzyme, consistent with the essentiality of Asp298. Global analysis of the slow UV/vis spectral changes observed during the reductive half-reaction of the D298A mutant with 2-phenylethylamine provided apparent rate constants for the formation and decay of TPQ(ssb) (k(obs) = 4.7 and 4.8 x 10(-4) s(-1), respectively), both of which are markedly smaller than those of the WT enzyme determined by rapid-scan stopped-flow analysis (k(obs) = 699 and 411 s(-1), respectively). Thus, Asp298 plays important roles not only in the alpha-proton abstraction from TPQ(ssb) but also in other steps in the reductive half-reaction. X-ray diffraction analyses of D298A crystals soaked with the substrate for 1 h and 1 week revealed the structures of TPQ(ssb) and TPQ(psb), respectively, as pre-assigned by single-crystal microspectrophotometry. Consistent with the stereospecificity of alpha-proton abstraction, the pro-S alpha-proton of TPQ(ssb) to be abstracted is positioned nearly perpendicularly to the plane formed by the Schiff-base imine double bond conjugating with the quinone ring of TPQ, so that the orbitals of sigma and pi electrons maximally overlap in the conjugate system. More intriguingly, the pro-S alpha proton of the substrate is released stereospecifically even in the reaction catalyzed by the base-lacking D298A mutant. On the basis of these results, we propose that the stereospecificity of alpha-proton abstraction is primarily determined by the conformation of TPQ(ssb), rather than the relative geometry of TPQ and the catalytic base.