Sir

It is not the case, as your News Feature1 states, that free-electron lasers (FELs) will have their “biggest impact” in biology. Such a prediction could normally be left to live or die peacefully, but the very high cost of X-ray FELs might make it an expensive funeral, and biological scientists should not have to pay the costs.

In structural biology, the fundamental limitation is radiation damage. This puts a ceiling on the dose of X-rays that can pass through a molecule before it is destroyed2. A small advantage might be gained from the use of pulsed X-ray sources such as the FEL promises to deliver, but this advantage will not be obtained in the first generation of X-ray FELs3 because the pulse lengths will be too long. A much bigger advantage has already been obtained during the 1990s by freezing protein crystals to liquid-nitrogen temperature and below. Together with improvements in the technology from expression and purification of proteins and brighter sources (third-generation synchrotrons), it has produced the current flood of new structures and has turned structural biology into the superbly productive field it is at present.

Sample freezing has also revolutionized electron microscopy, which can produce more structural information from a single protein molecule than can X-ray diffraction — before radiation damage terminates data collection. Electron cryomicroscopy already delivers many of the advantages hoped for from the X-ray FEL and has two further advantages. First, the cross-section for electron scattering is higher than for X-ray scattering, so working with single molecules is simpler. Second, by using lenses to focus the electrons, the diffracted beams can be collected and used to form an image which gives the phases of the Fourier components of the structure directly. X-ray diffraction gives only amplitudes.

The use of X-ray FELs in biology, as reported in your feature, is being hyped. Those who wish to build X-ray FELs should think of some good uses in another field and put aside for the moment the idea that they will have a big impact in biology. Your feature also suggests that there are relatively few good structural data on membrane proteins. While there is always more to be done, there are now excellent membrane protein structures for several members of at least 20 different membrane-protein families, obtained either by X-ray crystallography of three-dimensional crystals (mostly frozen) or by electron microscopy of frozen two-dimensional crystals. Progress towards atomic structures of membrane proteins without crystals will come from electron microscopy4, and nuclear magnetic resonance is also beginning to succeed with small membrane proteins5. I would not bet on getting any help from X-ray FELs.