Senior author

According to conventional wisdom, proteins are either completely folded or completely unfolded — the in-between steps all occur synchronously, making it impossible to examine them one by one. But Victor Muñoz and his colleagues at the University of Maryland, College Park, tested this idea with atom-by-atom analysis of protein folding. This was a great challenge, as the process involves protein purification, data capture using either nuclear magnetic resonance (NMR) or X-ray crystallography, and then a massive data crunch. Doing this for every step of the protein's folding process is costly and time-consuming. But, in doing so, this team found that sometimes the many atoms of a protein fold almost independently of one another, instead of all together (see page 317). Nature caught up with Muñoz to talk about how he set up the experiment.

Just how difficult was this work?

It was very time-consuming. The data initially looked a real mess. Instead of capturing one folding condition, we created many different conditions at different temperatures, and had to track hundreds of atomic signals that kept changing.

How did you know your approach was working?

When we looked at the atomic folding patterns and saw their complexity, it was what we had anticipated. But we still had our doubts. Was this complexity real, or were we looking at the wrong thing? Then we performed a beautiful control in which we obtained the conventional folding behaviour by simply averaging the hundreds of atomic signals in the protein. And when we saw the same thing in both low and high resolution, we were pretty sure we had it. It was very fulfilling and exciting.

What was different from what the conventional idea would have shown?

If proteins have to cross an energy barrier when they fold, the making of all the bonds that hold the atoms in their folded position would occur simultaneously, and therefore all atomic unfolding behaviours would be identical. What we saw is that pairs of atoms make bonds almost independently of the rest of the protein. It is the consolidation of the whole network of bonds that acts as a web, holding the folded structure together.

What's next?

We are developing theoretical predictions to determine which proteins show these properties. Small and helical proteins are more likely to conform to this type of behaviour, but we may find it is more ubiquitous than previously thought.