Introduction to crystallisation

Growing crystals of large macromolecules like proteins and DNA involves mixing a protein solution with crystallisation reagents then waiting to see if crystals form. Often extensive optimization is needed when an initial crystallisation condition "hit" is found, to obtain good quality crystals for data collection and structure solution.

Our standard method for screening for initial crystallisation conditions is called the sitting drop method. In this method a small stage sits above a reservoir containing the crystallisation reagents. Protein samples are concentrated to around 10 to 50 mg/ml, with suitable detergents in the case of membrane proteins. Small volumes (0.05 to 0.5 microlitres) of protein and well solution are mixed on the stage and the plate is sealed with tape. The plates are stored in a hotel either at 4oC or 20oC and images of the crystallisation drops are taken at intervals. With luck crystals grow as the system equilibrates.

For initial screening to find a starting condition with small crystals we use plates with 96 wells (Innovadyne/Wilden/MRC plates) and two protein stages on each of the 96 wells.We put 85 ul of crystallisation reagent into the well and 100 or 200 nanolitres of protein and crystallisation reagent onto the stage. Experiments often have to be repeated with a variety of volumes, concentrations, detergents and temperatures to find conditions in which crystals grow.

Once an initial condition is found we often scale up the crystallisation volumes, using 24 well Linbro plates and a hanging drop experiment. In this case 1 or more mmicrolitres of the protein and well solution are mixed on a siliconized cover slip, suspended over a well containing crystallisation reagents. In this case the well solutions can be mixed by a robot and the plates imaged by a robot, but the drops are set by hand.

Crystallisation screens

Crystallisation screens in 15 ml tubes (photo Paul Symes, Diamond)

The crystallisation reagents consist of precipitants (PEG, alcohols, salts), buffers, salts and additives. For initial trials we use commercial screens, in particular those designed by So Iwata's group for membrane proteins, although we do also use some screens for soluble proteins.

MemGold: Please note early versions on MemGold, before September 2007, had 10 different conditions complared to the more recent versions. The early version is referred to as MemGold beta, the current version is MemGold. The conditions for the current version of MemGold are in Newstead et al., in press.

Layout of drops in crystallisation plates

We use the layout shown below for crystallisation screens with 96 solutions. Where two 48 solution screens are combined, the second screen uses positions 49 to 96. Note: Structure screens 1 and 2 from Hampton have 50 solutions instead of 48. In this case we use solutions 1 to 48 of each screen, so 49 and 50 are not used.

Crystallisation robotics

The MPL crystallisation robotics Image © Diamond Light Source Ltd

We use the commercial screens described above, which have 96 different potential conditions included in each plate. We use robots to transfer the well solutions from 96 x 15 ml tubes in racks to 96-well plates. The plates are then sealed with PCR tape. We found this tape to give the best seal for storage of plates over several weeks. We prepare 10 plates in a batch using one set of disposible tips and we often do two or three batches in one day. These sealed plates are stored in an incubator at either 4 or 20 degrees C for several weeks and imaged at intervals.

When proteins samples are ready we take a selection of pre-prepared crystallisation plates, remove the tape and place a temporary lid on the plate. The plates are then placed on the deck of the Hamilton robot ready for use. Each plate in turn has the lid removed by the robot, the plate is then transferred to the cartesian nanodrop robot. This robot places a drop of well solution, between 100 and 200 nanolitres (0.1 to 0.2 microlitres) of well solution onto the stage and then add 100 to 200 nanolitres of protein to this drop. The plates are sealed with transparent tape using a plate sealer and then transferred to the hotel system.

At intervals specified by the user the hotel transfers the plate to a microscope which takes pictures of all the drops, either with brightfield or polarized light. These images are stored in a database. They can be inspected remotely either at Diamond or off site.

Recent Publication

The following publication describes the work of Simon Newstead, Sebastien Ferrandon and So Iwata in defining a new crystallisation screen for membrane proteins, called MemGold. This screen is sold by Molecular Dimensions Ltd.

"Rationalizing -helical membrane protein crystallization",
Newstead, S., Ferrandon, S. and Iwata, S.,
Protein Science, DOI: 10.1110/ps.073263108
http://www.proteinscience.org/cgi/content/abstract/ps.073263108v1