University of California, Berkeley, USA

A biochemist marvels at a molecule that shares his love of playing with fire.

I like to capture my students' attention by recounting how my early fascination with fire inspired my interest in the stability of sugars.

Glucose will 'burn' to carbon dioxide and water, liberating lots of energy. But it is stable enough that you can stamp on it without triggering the reaction — the energy barrier to the reaction is too high.

In my research, I am interested in how biology harnesses and controls oxygen reactivity. Most reactions, such as burning glucose, are held back by an energy barrier to getting things started. Enzymes can bypass this, finding a lower energy route through some reaction intermediate, to carry out a 'controlled burn'. Their control is not perfect, sometimes causing damage to both themselves and surrounding molecules, but by and large it works.

Typically, these enzymes have metal or organic components, which drive the oxidation. I often tell students that enzymes need their metal and organic cofactors because the 20 naturally occurring amino acids cannot carry out all the chemistry. Two recent papers shake that belief.

The surprise comes from the enzyme DpgC, which is involved in the biosynthesis of the antibiotic vancomycin. The first paper (C. C. Tseng et al. Chem. Biol. 11, 1195–1203; 2004) reports that DpgC uses oxygen in a complex dioxygenase reaction with no bound metal or organic cofactor.

More recently, researchers reported the structure of DpgC and confirmed that it has no cofactor (P. F. Widboom et al. Nature 447, 342–345; 2007). They find that the enzyme has a structure known as an oxyanion hole, which helps to stabilize the reaction intermediate.

I am still amazed that DpgC does oxygen chemistry with no help — and my students should be too.