Advances in High‐Throughput Crystallisation


Molecular biologists attempt to understand the inner workings of the cell, the smallest unit of all life forms on earth. The constituents that perform cellular functions are the proteins, enzymes and nucleotides. A full appreciation of the chemical and biological functions of the cellular constituents, and therefore life itself, requires knowledge of their three‐dimensional structures. X‐ray crystallography is among the most powerful methods to derive structural information from biological macromolecules. Before the three‐dimensional structure of biological macromolecules can be determined by X‐ray crystallography, they have to be assembled into the regular and periodic arrangements, which define a crystal. Biological macromolecules are irregularly shaped and inherently flexible, which makes their crystallisation very difficult. It frequently implies the testing of many different sample variations under many hundred conditions. This is the reason for the establishment of automated high‐throughput crystallisation facilities.

Key Concepts:

  • Progress in the life sciences requires knowledge of the cellular processes at a molecular level.

  • The principal components of the cell are proteins and nucleic acids.

  • The three‐dimensional structures of the cellular components define their functions.

  • X‐ray crystallography is the method of choice to determine the structure of biological macromolecules.

  • Crystallisation of biological macromolecules for X‐ray crystallography requires the testing of hundreds to thousands of conditions.

  • Automation supports many processes in molecular biology, by performing routine operations with the help of robotics.

  • High‐throughput crystallisation has significantly improved success rates in crystallography through the rapid and comprehensive screening of many sample constructs under many different conditions.

Keywords: structural biology; X‐ray crystallography; macromolecular crystallisation; molecular function; protein activity; sample characterisation

Figure 1.

Phase diagram for macromolecular crystallisation. The solubility line separates the soluble state of a mix of sample and precipitant from the region of insolubility. The metastable region designates the state, wherein precipitation will not occur in the absence of nuclei. Upon addition of nuclei, the metastable state would immediately crystallise around the nucleus, a phenomenon, which is exploited by microseeding (Bergfors, ). In the labile region, nuclei may form spontaneously and thus allow crystallisation. Beyond the nucleation zone the sample will precipitate as insoluble aggregates.

Figure 2.

Crystallisation methods. (a) In vapour diffusion, a small droplet of sample and precipitant (usually a 1:1 mix) is allowed to equilibrate in a closed container against a reservoir of pure precipitant solution. At the beginning of the experiment the vapour pressure of the droplet is higher than that of the reservoir, due to its lower precipitant concentration. Water transfer from the droplet into the reservoir leads to increasing sample and precipitant concentrations in the droplet, which may result in supersaturation. (b) For batch crystallisation sample and precipitant solution are mixed and sealed under inert oil. There is no change in the concentrations of either sample or precipitant. Batch crystallisation usually uses higher precipitant concentrations, having to start at supersaturation. (c) In free interface diffusion, sample and precipitant are combined in a reaction chamber (∼1 nL) and separated by a barrier. Removal of the barrier allows the precipitants to diffuse along the chamber and exposes the sample to a continuous gradient of precipitant concentrations. Counter diffusion experiments (not shown) work analogous to FID but in long (8–10 mm) capillaries, which produce shallower precipitant gradients and a finer sampling of the concentration range.

Figure 3.

Typical appearance of crystallisation droplets from VD experiments after equilibration. (a) The sample has not reached supersaturation and all components are still soluble. (b) The sample has precipitated by forming insoluble aggregates. (c) Microcrystals of ∼1–5 μm3, which are too small for current X‐ray sources. (d) Single crystal of about 80×60 μm2. Despite sufficient size, such crystals may require further improvement (see text).

Figure 4.

Integrated pipetting platform at EMBL Hamburg, consisting of a plate sealer (in blue) and a crystallisation and a pipetting robot for the set up of crystallisation plates and the production of crystallisation screens, respectively. The HydraII plusOne crystallisation robot is shown towards the right rear. The cycle time between two 96‐well crystallisation plates is 5 min. The Lissy2002 pipetting robot from Zinsser Analytic stores up to 120 different stock solutions in 50 and 15 mL falcon tubes (middle front). A sophisticated software controls the accurate pipetting of solutions into 96‐well deep well blocks.

Figure 5.

Visible (a) and UV image (b) of the same crystallisation drop. The objects in the drop are of an ambiguous nature. The UV image clearly shows that two objects on top contain protein, whereas two other similarly shaped objects do not.



Bergfors T (2003) Seeds to crystals. Journal of Structural Biology 142: 66–76.

Berry IM, Dym O, Esnouf RM et al. (2006) SPINE high‐throughput crystallization, crystal imaging and recognition techniques: current state, performance analysis, new technologies and future aspects. Acta Crystallographica Section D 62: 1137–1149.

Ericsson UB, Hallberg BM, Detitta GT, Dekker N and Nordlund P (2006) Thermofluor‐based stability of proteins. Analytical Biochemistry 357: 289–298.

McPherson A and Cudney B (2006) Searching for silver bullets: an alternative strategy for crystallizing macromolecules. Journal of Structural Biology 156: 387–406.

Rupp B (2003) Maximum‐likelihood crystallization. Journal of Structural Biology 142: 162–169.

Rupp B and Wang J (2004) Predictive models for protein crystallization. Methods 34: 390–407.

St. John FJ, Feng B and Pozharski E (2008) The role of bias in crystallization conditions in automated microseeding. Acta Crystallographica Section D 64: 1222–1227.

Tarendeau F, Boudet J, Guilligay D et al. (2007) Structure and nuclear import function of the C‐terminal domain of influenza virus polymerase PB2 subunit. Nature Structural & Molecular Biology 14: 229–233.

Watts D, Müller‐Dieckmann J, Tsakanova G, Lamzin VS and Groves MR (2010) Quantitative evaluation of macromolecular crystallization experiments using 1,8‐ANS fluorescence. Acta Crystallographica Section D 66: 901–908.

Yumerefendi H, Tarendeau F, Mas PJ and Hart DJ (2010) ESPRIT: an automated, library‐based method for mapping and soluble expression of protein domains from challenging targets. Journal of Structural Biology 172: 66–74.

Further Reading

Ducruix A and Giegé R (eds) (1992) Crystallization of Nucleic Acids and Proteins. New York: Oxford University Press.

McPherson A (2004) Introduction to protein crystallization. Methods 34: 254–265.

Contact Editor close
Submit a note to the editor about this article by filling in the form below.

* Required Field

How to Cite close
Meijers, Rob, and Mueller‐Dieckmann, Jochen(Jul 2011) Advances in High‐Throughput Crystallisation. In: eLS. John Wiley & Sons Ltd, Chichester. [doi: 10.1002/9780470015902.a0023171]