The HIS3+ gene of Saccharomyces cerevisiae was overexpressed in Escherichia coli and the recombinant imidazoleglycerol-phosphate dehydratase (IGPD) purified to homogeneity. Laser-desorption and electrospray m.s. indicated a molecular ion within 2 units of that expected (23833.3) on the basis of the protein sequence, with about half of the polypeptide lacking the N-terminal formylmethionine residue. IGPD initially purified as an apoprotein was catalytically inactive and mainly a trimer of M(r) 70,000. Addition of Mn2+ (but not Mg2+) caused this to assemble to an active (40 units/mg) enzyme (Mn-IGPD) comprising of 24 subunits (M(r) 573,000) and containing 1.35 +/- 0.1 Mn atoms/polypeptide subunit. An enzyme with an identical activity and metal content was also obtained when the fermenter growth medium of recombinant Escherichia coli was supplemented with MnCl2, and IGPD was purified through as Mn-IGPD rather than as the apoenzyme and assembled in vitro. Inhibition by EDTA indicated that the intrinsic Mn2+ was essential for activity. The retention of activity over time after dilution to very low concentrations of enzyme (< 20 nM) indicated that the metal remained in tight association with the protein. A novel continuous assay method was developed to facilitate the kinetic characterization of Mn-IGPD. At pH 7.0, the Km for IGP was 0.10 +/- 0.02 mM and the Ki value for inhibition by 1,2,4-triazole, 0.12 +/- 0.02 mM. In contrast with other reports, thiols had no influence on catalytic activity. The activity of Mn-IGPD varied with enzyme concentration in such a way as to suggest that it dissociates to a less active form at very low concentrations. Significant inhibition by the product, imidazole acetol phosphate, was inferred from the shape of the progress curve. Titration with, the potent competitive inhibitor, 2-hydroxy-3-(1,2,4-triazol-1-yl)propyl phosphonate indicated that Mn-IGPD contained 0.9 +/- 0.1 catalytic sites/protomer. The activity nearly doubled in the presence of high concentrations of Mn2+; the apparent Ks for stimulation was 20 microM. The basis of this effect was obscure, since there was no corresponding increase in the titre of active sites. Neither was there a discernable shift in the values of Km or Ki (above), although exogenous Mn2+ did reduce the optimum pH for kcat, from 7.2 to 6.8. On the basis of a single site/subunit, the maximum rate of catalytic turnover at 30 degrees C was 32 s-1.