Architectural Switching in Protein Evolution

It is useful to picture the collection of protein architectures in a particular cell or organism as the beautiful skyline of a huge city. Hundreds of different structures are present, but a given protein generally has a single fold, much in the way that a given building has a single unchanging architecture that makes up one part of a colorful view. The blueprints for assembling a given protein's fold correctly are encoded in the amino-acid sequence, which in turn is encoded in the nucleotide sequence of its gene. Like buildings, protein structures can be demolished or weakened by removing key architectural supports, such as might occur through mutations in the sequence. They may also tolerate some small architectural additions or minor modifications. But once a building is up, the basic architecture doesn't usually change radically over its lifetime, and traditional views of how protein structures behave and evolve are similarly conservative and static. If developers want to put up a more modern building on the space occupied by an old one, they usually knock it down and start over rather than try to twist what's already there into a new shape (unless it has some sentimental or historic value, like Boston's Fenway Park!). Similarly, it is thought to be fairly difficult to reshape existing proteins and still have them work correctly and maintain structural integrity.

But lately this analogy, while useful, has been under some strain. For one thing, a whole group of diseases, which include Alzheimer's and "Mad Cow", has been identified which involve "misfolding" events including radical architectural changes in proteins. It's also clear that some proteins, for example molecular motors or bacterial pore forming toxins, are incredibly structurally dynamic in the course of going about their normal business. Finally, the evolution of protein structure, once thought to be fairly uneventful, is showing itself to be more complex and dynamic than once realized.