Page 193 - Macromolecular Crystallography

182 MACROMOLECULAR CRYS TALLOGRAPHY

Edge Spectroscopy (XANES) can be used to esti- which uses an undulator as its X-ray source, stan-
mate the expected absorption energy of the anoma- dard exposure times are used for most samples. At
lous atoms. For example, SGX has determined the the selenium K edge, we typically use ∼1 sec for each
energy that corresponds to the maximum absorp- 1 degree oscillation. At the bromine K edge, where
tion of X-rays by pure selenomethionine. This value the X-ray intensity from the first harmonic of the
is used for all experiments that utilize selenome- undulator is approximately 40% of that at the Se
thionine to determine the required phases. With this edge, the exposure time is doubled. When smaller
approach, the limited X-ray lifetime of protein crys- or larger oscillations are used, exposure times are
talscanbedevotedtodeterminationofdiffraction. In adjusted proportionately. These values are appro-
addition to preserving crystals, this approach max- priate for datasets that include data from an angular
imizes the time devoted to data-collection in high- wedge up to 180 degrees. For anomalous dispersion
throughput mode. experiments, the amount of data collected generally
It should be noted that the absolute energy cal- increases by a factor of two. In these cases, exposure
ibrations of beamlines vary. Determining the peak times are typically reduced by a third, to preserve
position of an element within a protein relative to a sample integrity throughout the duration of data
known standard (such as the metallic form of that collection. Experience at SGX-CAT has shown that
element) can save time in future structural experi- this approach minimizes radiation damage while
ments. Accurate determination of absorption edge ensuring adequate diffraction intensity.
energies for various metals has been reported by Crystals with very strong diffraction may have an
Kraft et al. (1996). unacceptably high number of saturated reflections.
Asnotedabove, suchcrystalsareidentifiedautomat-
ically during the crystal evaluation process, based
12.6 Selection of exposure duration
on the average number of overloaded reflections on
The exposure time used for each diffraction image the screening images. For such crystals, the exposure
represents a compromise between maximizing the time is reduced. If a further decrease in the illumi-
intensity of the diffraction signal and minimizing nation of the sample is required, the intensity of the
crystal degradation during data collection. Nave incoming beam is adjusted.
and Garman have provided a brief review of the
current investigations on the sources of radiation 12.7 Sample tracking
damage (Nave and Garman, 2005). Henderson esti-
mated that the limiting X-ray dose for a protein When large numbers of crystals are examined at syn-
7
crystal is approximately 2×10 Grays (Gray = J/kg) chrotron beamlines, unambiguous identification of
(Henderson, 1990). Recentresultssuggestthatadose the samples is critical. The SPINE standard includes
50% higher than Henderson’s estimate can be tol- a unique barcode on the magnetic base for that pur-
erated by most protein crystals (Owen et al., 2006). pose. However, the SPINE protocol also assumes
The program RADDOSE can be used to estimate the that the bases will be reused. Hence, these 2D
maximum time that a crystal should remain in the barcodes alone do not permit unique identification
X-ray beam (Murray et al., 2004, 2005). Such calcu- of a given crystal.
lations rely on knowledge of both the characteristics SGX uses two bar codes to track samples. The first
of the beamline and the estimated composition of identifier is the 2D bar code on the pin bases, which
the unit cell. In addition, recent studies suggest that are used multiple times to mount individual crys-
there are significant variations in the X-ray dose that tals. The second identifier, a 1D bar code, is placed on
can be tolerated by the selenium in selenomethio- the vials in which crystals are shipped to the beam-
nine when determining diffraction phases through line. This barcode is unique to the crystal and, unlike
anomalous dispersion (Holton, 2007). the base, is not reused. The device used to read the
While it is possible to optimize exposure con- barcodes is shown in Fig. 12.3. This system requires
ditions for crystals individually, high-throughput that the 1D barcode, which links the physical crystal
efforts require an alternative approach. At SGX-CAT, with its description in the database, constitutes part