Reconstitution of membrane-integrated quinoprotein glucose dehydrogenase apoenzyme with PQQ and the holoenzyme's mechanism of action

Membrane-integrated quinoprotein glucose dehydrogenase from Acinetobacter calcoaceticus was produced by heterologous expression of the gene for it in an Escherichia coli recombinant strain. The apoenzyme (lacking the cofactor pyrroloquinoline quinone, PQQ) was solubilized with Triton X-100 and purified to homogeneity. Reconstitution of the apoenzyme to full activity in the assay was achieved with a stoichiometric amount of PQQ in the presence of Mg(2+). Just as for other PQQ-containing dehydrogenases where Ca(2+) fulfills this role, Mg(2+) anchors PQQ to the mGDH protein and activates the bound cofactor. This occurs in a precise way since high anomer specificity was found for the enzyme toward the sugars tested. Although the steady-state-type kinetics were as expected for a dye-linked dehydrogenase (ping-pong) and the PQQ in it was present in oxidized form, addition of glucose to the holoenzyme resulted in a very slow but continuous production of gluconolactone, i.e., the reaction did not stop after one turnover, with O(2) apparently acting as an (albeit poor) electron acceptor by reoxidizing PQQH(2) in the enzyme. The surprisingly low reactivity with glucose, in the absence of dye, as compared to the activity observed in the steady-state assay appeared to be due to formation of an anomalous enzyme form, mGDH*. Formation of normal holoenzyme, mGDH, reducing added glucose immediately to gluconolactone (in one turnover), was achieved by treating mGDH* with sulfite, by reconstituting apoenzyme with PQQ in the presence of sulfite, or by applying assay conditions to mGDH* (addition of PMS/DCPIP). As compared to other quinoprotein dehydrogenases, mGDH appears to be unique with respect to the mode of PQQ-binding, as expressed by the special conditions for reconstitution and the absorption spectra of the bound cofactor, and the reactivity of the reduced enzyme toward O(2) The primary cause for this seems not to be related to a different preference for the activating bivalent metal ion but to the special way of binding of PQQ to mGDH.