Xray Crystallography/Radiation Damage
The problem of radiation damage to crystals, even to cryocooled crystals, has become a major problem again since the advent of the brilliant beam 3rd generation synchrotrons.
Basis of Radiation Damage
It has long been known that radiation damage to proteins is important, in 1962 Blake and Phillips showed that damage was proportional to the dose received by the crystal and that each 8keV photon was capable of disrupting 70 molecules and disordering 90 more.
At 12.7 keV, of the 2% of X-rays that interact with the crystal, 84% interact through the photo-electric effect (Murray J and Rudiño-Piñera E, 2005). Each X-ray photon can only produce one photoelectron, however each photoelectron can result in over 500 secondary electrons of a lower energy (O'Neil et al. 2002).
Even at cryo temperatures electrons are still mobile within the protein, they can exploit electron tunnelling effects to jump along the protein backbone chain (Jones et al. 1987).
Types of Radiation Damage
Primary radiation damage is dose dependent and caused by the interactions X-ray beam photons creating photoelectric electrons. This is an inevitable consequence of X-ray crystallography.
Photoelectric electrons from primary radiation damage can then go on to further damage the protein through the creation of free radical, this process is termed secondary radiation damage, in a manner which is both time and temperature dependent. The diffusion of free radicals through the crystal can be slowed by cryocooling.
Direct / Indirect
Both primary and secondary radiation damage can be either direct, directly altering the protein, or indirect, altering the surrounding solvent. Indirect effects can still be just as damaging to the protein, the radiolysis of water can create OH, H, H+ and hydrated electrons which are especially destructive.
Damage occurs to proteins in a specific manner. The ability of free electrons to tunnel along the peptide backbone (Jones et al., 1987) provides a mechanism for specific structural damage to occur. Structural damage occurs in the order of covalent bond strength (Garman and Owen, 2006);
- Disulphide bond breakage.
- Decarboxylation of aspartates and glutamates.
- Loss of OH groups from tyrosines.
- C-S bond cleavage in methionines.
- (Burmeister, 2000; Ravelli and McSweeny, 2000; Weik et al, 2000)
Damage also occurs to metal centres within proteins (Carugo and Carugo, 2005) causing reduction. This can create a problem if most of the pdb structures contain reduced (or partially reduced) metal centres. Active sites of proteins are also susceptible to damage as with the FAD cofactor in DNA apophotolyase (Kort et al., 2004).
It is also known that specific damage occurs before the diffraction pattern is visibly compromised (Ravelli and McSweeny, 2000).
- Change in unit cell dimensions.
- Increase Wilson B values.
- Decreased diffraction power of crystal.
- Loss of high resolution data.
A change in the unit cell dimensions is also characteristic of radiation damage (Ravelli and McSweeny, 2000), however they cannot be used as a measure of damage due to the wide variations in change seen even for crystals of the same protein (Murray and Garman, 2002). Weik et al. (2001) showed a dependence of unit cell volume with temperature, however temperature effects are not thought to be significant to radiation damage in beamlines (Nicholson et al., 2001; Kuzay et al., 2001). Unit cell volume changes are important when you consider that a 1/2% change in all cell dimensions gives a 15% change in general reflection intensities (Crick and Magdoff, 1957), especially when you are only trying to measure a 6-10% change in intensities for anomalous dispersion experiments (Hendrickson and Ogate, 1997).
Rate of Damage
damage proportional to dose
cryo room temp