crystal and subsequently determine the three-dimensional structure
of the RNA–protein complex if the crystal diffracts to a reasonable
resolution (preferably around 2.5 A ˚ or higher). However, more
commonly crystals do not diffract sufficiently well enough to collect
a complete data set or are otherwise of poor quality, e.g., too small,
or the crystals are multiple or split. In such cases, the preliminary
coarse screen condition(s) must be fine-tuned by varying each
crystallization reagent in order to attempt to improve crystal quality. Typically, the protein concentration, RNA:protein ratio, precipitant concentration, additive concentration, buffer pH, and
temperature are varied. When several conditions are identified in
the initial coarse screens that produce crystals with similar morphology, it is useful to cross-correlate conditions, e.g., pH, precipitant,
or additives, to inform the design of follow-up fine (optimization)
screens. Alternatively, if crystals with different morphologies are
observed, the conditions should be treated independently since
the resultant crystals may have different diffraction properties.
Liquid handling robots such as the Dragonfly
® (SPT Labtech)
simplify the process of designing and dispensing optimization
screens into 96-well SBS format plates and can dramatically speed
up the optimization process.
3.12 Crystal
Harvesting, Freezing,
and Data Collection
When crystals are observed in the crystallization experiments, it is
advisable to seek assistance from a colleague with a crystallographic
background to aid crystal mounting and freezing since protein
crystals tend to be fragile and can be difficult to handle and manipulate. Crystals must be stabilized and cryoprotected, typically using
the reservoir solution supplemented with cryoprotectant, e.g., 20%
ethylene glycol or glycerol, prior to harvesting in a cryo loop and
flash freezing in liquid nitrogen (see Note 7). Frozen crystals can be
stored long term in a liquid nitrogen dewar in vials or unipucks (see
Note 8), and screened either on an in-house X-ray source or at a
synchrotron beamline to evaluate diffraction quality. Typically, data
extending to 2.5 A ˚ resolution or higher are desirable for novel
RNA–protein complexes so that a three-dimensional model for
the complex can be accurately fitted into the electron density
map. However, in more challenging cases, data at 3 A ˚ resolution
or lower may be sufficient to fit the overall fold of the protein and
RNA. Data acquisition involves recording a series of X-ray diffraction images using detectors such as the Rigaku Saturn 944+ charge
coupled device (CCD) and Dectris Eiger, which are commonly
found on in-house and synchrotron sources, respectively.
In-house data can be integrated and scaled to produce an MTZ
file, for example, using the StructureStudio™ software platform for
data collected on Rigaku systems or, alternatively, MOSFLM [20]
and the CCP4 suite of programs (see Note 4) and its graphical user
interface [5, 21]. Synchrotron data are usually processed on-the-fly
using Xia2 [22], which is incorporated into the ISPyB (Information
System for Protein CrystallographY Beamlines), LIMS (laboratory
436
Andrew P. Turnbull and Xiaoqiu Wu
of the RNA–protein complex if the crystal diffracts to a reasonable
resolution (preferably around 2.5 A ˚ or higher). However, more
commonly crystals do not diffract sufficiently well enough to collect
a complete data set or are otherwise of poor quality, e.g., too small,
or the crystals are multiple or split. In such cases, the preliminary
coarse screen condition(s) must be fine-tuned by varying each
crystallization reagent in order to attempt to improve crystal quality. Typically, the protein concentration, RNA:protein ratio, precipitant concentration, additive concentration, buffer pH, and
temperature are varied. When several conditions are identified in
the initial coarse screens that produce crystals with similar morphology, it is useful to cross-correlate conditions, e.g., pH, precipitant,
or additives, to inform the design of follow-up fine (optimization)
screens. Alternatively, if crystals with different morphologies are
observed, the conditions should be treated independently since
the resultant crystals may have different diffraction properties.
Liquid handling robots such as the Dragonfly
® (SPT Labtech)
simplify the process of designing and dispensing optimization
screens into 96-well SBS format plates and can dramatically speed
up the optimization process.
3.12 Crystal
Harvesting, Freezing,
and Data Collection
When crystals are observed in the crystallization experiments, it is
advisable to seek assistance from a colleague with a crystallographic
background to aid crystal mounting and freezing since protein
crystals tend to be fragile and can be difficult to handle and manipulate. Crystals must be stabilized and cryoprotected, typically using
the reservoir solution supplemented with cryoprotectant, e.g., 20%
ethylene glycol or glycerol, prior to harvesting in a cryo loop and
flash freezing in liquid nitrogen (see Note 7). Frozen crystals can be
stored long term in a liquid nitrogen dewar in vials or unipucks (see
Note 8), and screened either on an in-house X-ray source or at a
synchrotron beamline to evaluate diffraction quality. Typically, data
extending to 2.5 A ˚ resolution or higher are desirable for novel
RNA–protein complexes so that a three-dimensional model for
the complex can be accurately fitted into the electron density
map. However, in more challenging cases, data at 3 A ˚ resolution
or lower may be sufficient to fit the overall fold of the protein and
RNA. Data acquisition involves recording a series of X-ray diffraction images using detectors such as the Rigaku Saturn 944+ charge
coupled device (CCD) and Dectris Eiger, which are commonly
found on in-house and synchrotron sources, respectively.
In-house data can be integrated and scaled to produce an MTZ
file, for example, using the StructureStudio™ software platform for
data collected on Rigaku systems or, alternatively, MOSFLM [20]
and the CCP4 suite of programs (see Note 4) and its graphical user
interface [5, 21]. Synchrotron data are usually processed on-the-fly
using Xia2 [22], which is incorporated into the ISPyB (Information
System for Protein CrystallographY Beamlines), LIMS (laboratory
436
Andrew P. Turnbull and Xiaoqiu Wu