By tuning the tape's speed, the researchers controlled how long each crystal reacted before an XFEL pulse struck it, freezing an atomic-resolution image. Stacking thousands of those snapshots yielded a frame-by-frame "molecular movie" of penicillin biosynthesis.

The approach exposed two key stages:

• A thioaldehyde intermediate that forms immediately before the β-lactam ring closes.
• A monocyclic β-lactam — the first ring-shaped structure en route to the full penicillin scaffold.

Both intermediates had been too difficult to observe directly before.

Why do these steps matter?

The team also found that water molecules tucked inside the enzyme's active site guide the reaction, and that subtle movements of the entire protein help steer each chemical step. That means enzyme shape and enzyme chemistry cooperate in building penicillin — a principle the researchers say could guide future engineering.

Dr. Patrick Rabe, first author and a Wellcome Career Development Award investigator at Oxford, said: "These insights matter because they can inform how we think about antibiotic biosynthesis and future strategies for developing antibacterial medicines. By understanding this process in atomic detail, we can begin to think about engineering these enzymes to produce new or improved antibiotic scaffolds."

What does Oxford's history add to the story?

Oxford ties to penicillin run deep. Howard Florey, Ernst Chain and colleagues turned penicillin into a usable drug during World War II. In 1945, Nobel-winning chemist Dorothy Hodgkin mapped penicillin's structure with X-ray crystallography — the same basic technique this new study extends to a working enzyme on the move.

Today's result effectively picks up Hodgkin's thread eight decades on.

Does this reach beyond antibiotics?

Yes. IPNS belongs to a large family of iron-dependent oxygenases that also help humans sense oxygen levels. The mechanistic principles the team uncovered — how water, iron and protein motion cooperate — could shape enzyme engineering and catalyst design well outside antibiotic research.

What about antimicrobial resistance?

Roughly one in six bacterial infections globally already resists existing antibiotics. The World Health Organization and others warn that the shrinking pipeline threatens cancer care, transplant surgery and routine infection treatment.

Schofield connected the basic chemistry to that clinical crisis: "As rates of antimicrobial resistance continue to rise, understanding this process will ultimately help us make existing antibiotics more efficient and design new antibiotic structures."

Who was involved?

The project united structural biologists, spectroscopists, computational chemists and engineers from Oxford, Diamond Light Source, Lawrence Berkeley National Laboratory, PAL-XFEL and SLAC National Accelerator Laboratory.

Dr. Allen M. Orville, group leader of the XFEL Hub at Diamond, said: "The results show the power of combining complementary expertise and X-ray techniques to move beyond static structures and reveal how enzymes work in real time."

A note on limitations

The work is preliminary in the sense that only IPNS has been filmed this way so far. Translating the mechanistic insight into marketable antibiotics will require further engineering of IPNS and related enzymes, and the technique applies only to enzymes that crystallize and tolerate oxygen-triggered reactions.

The paper, "Unanticipated intermediates during isopenicillin N synthase catalysis identified by time-resolved X-ray free-electron laser studies," appeared in Nature Catalysis (2026). DOI: 10.1038/s41929-026-01618-4.