A roundabout way to obtain protein structural information

In the paper "Rational design of a structural and functional nitric oxide reductase" Yeung et al. describe a way to get structural information on an enzyme, nitric oxide reductase, which could not be crystallised. They mutated sperm whale myoglobin to yield the same enzymatic activity and then crystallised that... Not as nice as crystallising the real thing of course, but if it's the only thing that works...