the observed conditions. While keeping the buffer (trisodium
citrate at 100 mM) and ethylene glycol (1% (v/v)) concentra-
tions fixed, a two-dimensional screening is done by systemati-
cally varying ammonium sulfate and lithium sulfate
concentrations between 250 mM and 1 M. Such a broad
crystallization range helps to ensure that crystals are obtained
from different batches of proteins which could be slightly
different from each other.- AMPK crystals cannot be stored indefinitely. In our experience,
after ~14 days, crystals are less amenable to handling, often
associated with a film or precipitate, and yield poorer
diffraction data. - Minimize the time involved in looping or manipulating crys-
tals. As the crystallization drop is exposed to air, salt crystals
will begin to form and complicate harvesting. - Ensure AMP is maintained throughout soaks and cryo-
protection. Removal of AMP during crystal manipulation has
yielded poorer diffraction. - To confirm ligand binding pose in the 3–3.5 A ̊ resolution
regime, we have found it helpful to determine the structure
with halogenated analogs of ligands (e.g., Br-containing) to
enable collection of anomalous diffraction data to uniquely
identify the location of electron-dense bromine atom [9]. - The direct output of X-ray diffraction from protein crystals is
not an automated molecular structure. Rather, the observed
diffraction data is used to compose electron density distribu-
tion within the crystals which a crystallographer interprets in
the form of an atomic model. As such, a “crystal structure” is a
crystallographer’s interpretation of experimental electron den-
sity which best explains the observed diffraction data while
conforming to physical and chemical restraints. Given the
dynamic nature of AMPK, the resolution range for full-length
AMPK structures is approximately 3–3.5 A ̊, a threshold that is
slightly low for typical structure-based drug design. As such,
particular care must be taken in both modeling and interpret-
ing ligand pose. - The final AMPK enzyme concentration used in the assay
should be predetermined to ensure the reaction is in the linear
phase during the selected time course. It is recommended to
perform a time course of the reaction at various enzyme con-
centrations to select a final condition where approximately less
than 10% of the substrate has been converted to product. - The SAMS peptide and ATPKmvalues vary depending on the
AMPK isoform. Use the reportedKmfor the desired AMPK
isoform in choosing the range of substrate concentrations. For
initial experiments, we typically use just a few substrate
52 Ravi G. Kurumbail et al.