Incubation of a 1 : 1 mixture of b-(L-a-aminoadipoyl)-L-cysteinybvalinet (L,L,D-ACV) and L , L , D -A [ ~, ~-~H ~] C V with a cell-free extract of isopenicillin N synthetase from Cephalosporium acremonium results in preferential conversion of the fully protiated substrate into isopenicillin N; alternatively a similar experiment with a 1 : 1 mixture of L,L,D,-ACV and L,L,D-AC[~-~H]V gave isopenicillin N without isotopic discrimination between the two substrates.

Isopenicillin N synthetase was extracted from Cephalosporium acremonium and purified about 200-fold. The product showed one major protein band, coinciding with synthetase activity, when subjected to electrophoresis in polyacrylamide gel. An isopenicillin N synthetase from Penicillium chrysogenum was purified about 70-fold by similar procedures. The two enzymes resemble each other closely in their Mr, in their mobility on electrophoresis in polyacrylamide gel and in their requirement for Fe2+ and ascorbate for maximum activity. Preliminary experiments have shown that a similar isopenicillin N synthetase can be extracted from Streptomyces clavuligerus.

Deacetoxycephalosporin C synthetase (expandase) from Cephalosporium acremonium (Acremonium chrysogenum) was purified to near homogeneity as judged by SDS/polyacrylamide-gel electrophoresis. The enzyme (Mr about 40,000) exhibited a pH optimum around 7.5. It required 2-oxoglutarate (Km 0.04 mM), Fe2+ and O2 as cofactors, and ascorbate and dithiothreitol were necessary for maximum activity. It was stable for over 4 weeks at -70 degrees C in the presence of 1 mM-dithiothreitol. Activity was inhibited by the thiol-quenching reagent N-ethylmaleimide, the metal-ion-chelating reagent bathophenanthroline, and NH4HCO3. The highly purified enzyme also showed deacetoxycephalosporin C hydroxylase (deacetylcephalosporin C synthetase) activity, indicating that both expandase and hydroxylase activities are properties of a single protein. These activities could not be separated by ion-exchange, dye-ligand, gel-filtration or hydrophobic chromatography. A beta-sulphoxide and a 3 beta-methylene hydroxy analogue of penicillin N were synthesized to test as potential intermediates in the ring-expansion reaction, Neither compound was a substrate for the enzyme. A synthetic analogue in which the 3 beta-methyl group and the 2-hydrogen atom of penicillin N were replaced by a cyclopropane ring was not a substrate but was a reversible inhibitor of the enzyme.

Cyclisation of the cyclopropyl-containing substrates ( 6 ) and ( 7) by the enzyme isopenicillin N synthetase (IPNS) gave the bicyclic (3-lactams (9) and (lo), respectively; these products support the hypothesis that a carbon-centred free radical or equivalent iron-carbon bonded intermediate is involved in the carbon-sulphur bond formation.