How does radiation damage in protein crystals depend on X-ray dose?

Is radiation damage to cryopreserved protein crystals strictly proportional to accumulated dose at the high-flux density of beams from undulators at third-generation synchrotron sources? The answer is "yes," for overall damage to several different kinds of protein crystals at flux densities up to 10(15) ph/sec/mm(2) (APS beamline 19-ID). We find that, at 12 keV (1 A wavelength), about ten absorbed photons are sufficient to "kill" a unit cell. As this corresponds to about one elastically scattered photon, each unit cell can contribute only about one photon to total Bragg diffraction. The smallest crystal that can yield a full data set to 3.5 A resolution has a diameter of about 20 microm (100 A unit cell).