Oxford researchers film the fleeting steps an enzyme takes to build penicillin's ring

Oxygen starts the reaction. The team grew the enzyme as microcrystals without oxygen, put thousands of droplets of them on a moving tape 2 mm wide, and ran the tape into an oxygen-filled chamber ^[5]. Oxygen diffused into the crystals quickly, so the reaction began across the whole sample at the same moment ^[5].

Tape speed set the clock. The speed of the tape fixed how long each crystal reacted before it reached the point where an ultrafast X-ray free-electron laser pulse took an atomic-resolution snapshot ^[5]. The team assembled thousands of snapshots, each taken at a known delay, into a frame-by-frame sequence. It shows species that exist for fractions of a second, at physiological temperature and pressure ^[6].

The field had mostly relied on static crystal structures. This experiment instead followed the reaction as it ran ^[13]. Static crystallography has a long history with this molecule: the Oxford chemist Dorothy Hodgkin used it to solve penicillin's structure in 1945 ^[12]. For decades, though, the short-lived intermediates between the linear peptide and the finished ring were too hard to observe directly ^[8]. A time series lets the steps in between be seen instead of inferred from the two ends. The results appear in Nature Catalysis ^[1].

I would hold the water result most loosely. Time-resolved snapshots can show where water molecules sit at each stage of the reaction ^[4], but that tells you where the water is during the chemistry. Before calling it a guide, I'd want to see the reaction slow or change course when that water is displaced.

The release links the work to rising antimicrobial resistance and to the small number of new antibiotics in development. It calls understanding how nature builds these molecules an important step toward replenishing the pipeline ^[10]. The beta-lactam ring is a strained structure that interferes with bacterial cell wall synthesis until the wall fails ^[9]. The release also connects the findings to enzyme engineering as well as drug development ^[14]. Engineering is the shorter route. Anyone redesigning IPNS to make altered ring systems would start from a step-by-step account of what the enzyme does. The thing this doesn't tell you is how close such a redesign is. The release does not describe a candidate compound or a programme built on the result.