The coordination number of the catalytic zinc ion in alcohol dehydrogenase

An old controversial issue in the reaction mechanism of alcohol dehydrogenase is whether the catalytic zinc ion is four-coordinate in all steps (as crystallography indicates) or if some complexes are five-coordinate (as some spectroscopic studies suggest). This has been studied by various theoretical methods. First, I performed quantum chemical geometry optimisations on a number of realistic models of the active site in alcohol dehydrogenase, viz. Zn(SH)2XL(H2O)0-1, where X is NH3 or imidazole, and L is H2O, CH3 OH, CH3 CH2OH, or the corresponding anions and aldehydes [15]. These studies showed that a zinc ion in vacuum with the same ligands as in alcohol dehydrogenase prefers four- before five-coordination by about 20 kJ/mole. With negatively charged L-ligands, a five-coordinate minimum does not even exist.

These structures were later used to construct a classical force-field parameterisation of four- and five-coordinate zinc ions, tailored for alcohol dehydrogenase [16]. The parameterisation was used in molecular dynamics and molecular mechanics simulations of the active site zinc ion in alcohol dehy-drogenase, which indicated that in the enzyme, a four-coordinate active-site zinc ion is at least 37 kJ/mole more stable than a five-coordinate one. Finally, I constructed a combined quantum chemical and molecular mechanical geometry optimisation program (COMQUM, see below) and showed that five-coordinate zinc complexes in alcohol dehydrogenase are disfavoured not only by the zinc ligands (by about 20 kJ/mole) but also by the enzyme (about 80 kJ/mole) [20]. The total energy difference is so large that five-coordinate complexes may hardly play any significant role in the reaction mechanism of the enzyme. Encouragingly, this conclusion was later confirmed by two experimental publications that reinterpret most spectroscopic observations of five-coordinate complexes in alcohol dehydrogenase [Inorg. Chem. 34 (1995) 218; Biochemistry 34 (1995) 7145].

The study of the zinc coordination number in alcohol dehydrogenase also resulted in two papers on related subjects. First, I showed that Glu-68, a conserved amino acid located 0.5 nm from the catalytic zinc ion in the crystal structure, may intermittently coordinate to the zinc ion [17] (this has later been confirmed, see below). The active site seems actually to be constructed to enable such a binding. Glu-68 probably facilitate the exchange of zinc ligands. Second, I considered the structural  zinc ion [19] and show that all its four cysteine ligands are deprotonated and not only two that has previously been suggested. It turned out that the Zn-S bond distance is very sensitive to the theoretical treatment. The experimental distance can only be reproduced if both electronic correlation and a detailed picture of the surrounding enzyme are considered (a continuum model does not work). This was a triumph for COMQUM.