ARTICLE IN PRESS

Progress in Biophysics and Molecular Biology 88 (2005) 329–337 www.elsevier.com/locate/pbiomolbio

Review

Methods for separating nucleation and growth in protein crystallisation Naomi E. Chayen Biological Structure and Function Section, Division of Biomedical Sciences, Sir Alexander Fleming Building, Imperial College London, Exhibition Road, London SW7 2AZ, UK Available online 22 September 2004

Abstract The availability of high-quality crystals is crucial to the structure determination of proteins by X-ray diffraction. With the advent of structural genomics the pressure to produce crystals is greater than ever before. Finding favourable conditions for crystallisation is usually achieved by screening of the protein solution with numerous crystallising agents. Optimisation of the crystallisation conditions involves the manipulation of the crystallisation phase diagram with the aim of leading crystal growth in the direction that will produce the desired results. This article highlights recent advances in experimental methods for improving crystal size and quality by separating the nucleation and growth phases of crystallisation using the vapour diffusion and microbatch techniques. r 2004 Elsevier Ltd. All rights reserved. Keywords: Microbatch; Nucleation; Phase diagrams; Protein crystallisation; Vapour diffusion

Contents 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 1.1. The crystallisation phase diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330

2.

Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 2.1. Separation of nucleation and growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 Tel.: +44-20-759-432-40; fax: +44-20-759-431-69.

E-mail address: [email protected] (N.E. Chayen). 0079-6107/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.pbiomolbio.2004.07.007

ARTICLE IN PRESS N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

330 2.2. 2.3. 2.4. 3.

Solubility versus super-solubility curves of the phase diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Dilution techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 The application of light-scattering techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

Summary and future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336

1. Introduction 1.1. The crystallisation phase diagram The crystallisation process can be illustrated by a phase diagram that shows which state (liquid, crystalline or amorphous solid [precipitate]) is stable under a variety of crystallisation parameters. It provides a means of quantifying the influence of the parameters such as the concentrations of protein, precipitant(s), additive(s), pH and temperature on the production of crystals. Hence phase diagrams form the basis for the design of crystal growth conditions (Ducruix and Giege, 1992; McPherson, 1999; Chayen et al., 1996 and references therein). Crystallisation proceeds in two phases, nucleation and growth. Nucleation which is a prerequisite for crystallisation requires different conditions than those of growth. Once the nucleus has formed, growth follows spontaneously (Ataka, 1993; Ducruix and Giege, 1992; McPherson, 1999). Fig 1 shows an example of a typical crystallisation phase diagram. The figure schematically illustrates four areas: (i) an area of very high supersaturation where the protein will precipitate; (ii) an area of moderate supersaturation where spontaneous nucleation will take place; (iii) an area of lower supersaturation just below the nucleation zone where crystals are stable and may grow but no further nucleation will take place (this area is referred to as the metastable zone which is thought to contain the best conditions for growth of large well-ordered crystals); (iv) an undersaturated area where the protein is fully dissolved and will never crystallise (Chayen et al., 1996). In an ideal experiment, once nuclei have formed, the concentration of protein in the solute will drop, thereby naturally leading the system into the metastable zone where growth should occur, without the formation of further nuclei (Ducruix and Giege, 1992; McPherson, 1999; Bergfors, 1999). However, the ideal experiment does not often happen and more often than not, excess nucleation occurs, resulting in the formation of numerous low-quality crystals. The aim is therefore to devise methods that will enable the experimenter to lead the experiment from nucleation to growth in order to ensure the desired results.

2. Methods 2.1. Separation of nucleation and growth Crystal growth can be controlled by separating the phases of nucleation and growth. This can be achieved by bypassing the nucleation zone by inserting crystals, crystal seeds or other nucleants

ARTICLE IN PRESS N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

331

Fig. 1. Schematic illustration of a protein crystallisation phase diagram.

directly into the metastable zone. Many successful examples of such experiments have been reported over the years (e.g. McPherson and Schlichta, 1988; Stura, 1999; Chayen et al., 2001; Bergfors, 2003). An alternative approach is to initiate the crystallisation process at conditions that induce nucleation and subsequently lead the system to metastable conditions which promote optimal growth. This article highlights some recent advances to the latter approach. 2.2. Solubility versus super-solubility curves of the phase diagram Although phase diagrams offer one of the basic and most important pieces of knowledge necessary for growing protein crystals in a rational way, they are not often employed because accurate quantitative phase diagrams require numerous solubility measurements. (The solubility is defined as the concentration of protein in the solute that is in equilibrium with crystals). Reaching equilibrium can take several months (Ataka, 1993 and references therein) because proteins diffuse slowly. An additional limiting factor is that solubility measurements require large amount of sample. A method for pinpointing the metastable zone has been devised (Wilson, 2003), but the procedure is still too complicated for immediate application to routine use. The area of conditions called the ‘metastable zone’ is situated between the solubility and supersolubility curves on the crystallisation phase diagram (Fig. 1). The supersolubility curve is defined as the line that separates the conditions where spontaneous nucleation (or phase separation or precipitation) occurs, from those where the crystallisation solution remains clear if left undisturbed (Ducruix and Geige, 1992). The supersolubility curve is less well-defined than the solubility curve but experimentally, it is found to a reasonable approximation, much more easily. It has been reported that for practical

ARTICLE IN PRESS 332

N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

purposes, it is sufficient to obtain the supersolubility curve. To construct it, one must set up crystallisation trials, varying at least two conditions (one of which is typically the protein concentration) and plot their outcomes on a two-dimensional parameter grid. The supersolubility curve can be obtained rapidly using robots and can aid in separation of nucleation and growth using seeding and other means. A diagram containing the supersolubility curve (and not the solubility curve) is called a ‘working phase diagram’ (Saridakis and Chayen, 2003). The first ‘working phase diagram’ which was generated for the purpose of seeding was reported by Korkhin et al. (1995). The authors constructed such a phase diagram in approximately 4 h using the automated microbatch technique to dispense six protein concentration values versus 24 different precipitant concentrations. Conditions for microseeding were predicted from that phase diagram and subsequent to seeding, crystals of an alcohol dehydrogenase diffracting to 2 A˚ were grown reproducibly in 2 ml drops. Nowadays such experiments could easily be performed using smaller volume drops. More recently, similar phase diagrams were constructed automatically in order to find conditions for successfully inserting external nucleants (Chayen et al., 2001). Although seeding is very successful and it is the most widely used method for separation of nucleation and growth (Stura, 1999 and references therein), seeding often involves handling of fragile crystals, and other than streaking, seeding manoeuvres are not very simple to perform. Insertion of external nucleants is invasive and can give rise to disturbance in the trial drops. A non-invasive simple means of separating nucleation and growth by changing the temperature at which trials are incubated has been successfully applied for several proteins (e.g. Rosenberger, 1993; Haire Llyod, 1996). However, the use of temperature variation during the experiments is limited because not all proteins are sensitive to temperature change. 2.3. Dilution techniques An alternative means of conducting trials with a similar outcome to seeding but without the need to handle crystals or nucleants, is by dilution. An additional benefit of dilution methods is that they are more amenable to automation compared with seeding techniques. A very common occurrence in crystallisation is the formation of clusters of non- diffracting crystals that can not be improved by merely changing the crystallisation parameters. In such cases, a working phase diagram that provides the supersolubility curve can be generated (either manually or using a robot). In the case of hanging drops, the coverslips holding the drops are incubated for some time over reservoir solutions that normally give many small crystals. After a given time (before crystals are visible), the coverslips are transferred over reservoirs containing lower precipitant concentrations that would normally yield clear drops (Fig. 2). The time of transfer is selected by reference to the time in which it took to see the first crystals in the initial screens. For example, if crystals appeared within 24 h, nucleation would have occurred anytime between setting-up the experiments to several hours before the crystals appear. Hence the transfer should be done at intervals of 1–2 h after set-up. Trials that are transferred too soon will produce clear drops while those that are transferred too late will yield low-quality crystals. The transfer lasts just 1–2 s. This technique has produced significant improvement in crystal order of a number of proteins (e.g. Saridakis and Chayen, 2000; Saridakis and Chayen, 2003). Fig. 3 shows crystals of C-phycocyanin obtained by

ARTICLE IN PRESS N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

333

Fig. 2. Transfer of a coverslip holding a crystallisation drop from nucleation to growth conditions.

Fig. 3. Crystals of C-phycocyanin diffracting to 1.45 A˚ grown in a hanging drop that was transferred from nucleation to metastable conditions. Crystal dimensions: 200  200  300 mm.

such a transfer experiment. The crystals reproducibly diffracted to 1.45 A˚ compared to 2 A˚ using standard optimisation techniques. The crystal structure has been solved by Nield et al. (2003). The experiments described above were performed using the standard Linbro plates in which the reservoirs are sealed with grease or oil. This makes it difficult to remove the cover-slip after a period of time and to be certain of the effectiveness of the seal over the new reservoir. The recently developed Nextal Crystallisation Tool (Nextal Technologies, Canada) consists of wells that are sealed by a screw cap that incorporates the cover-slip. The screw cap design affords easy transfer of drops from one reservoir to another making this tool particularly useful for setting up trials for decoupling the phases of nucleation and growth. The Nextal crystallisation tool has been further developed to include a new feature that enables the induction and subsequent arrest of nucleation in a quantitative and reproducible way. The

ARTICLE IN PRESS 334

N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

outer circumference of each well has been marked with graduations (Fig. 4) so that the extent of the seal can be accurately set for each individual reservoir. The graduations on the well are used to adjust the tightness of the seal in order to allow variable amounts of evaporation without exposing the drop, thus improving on previous experiments in which the drops were exposed (e.g. Jovine, 2000). The graduations marks also ensure reproducibility of the trials. The experiments involve loosening of the screw caps above the wells at the time of setting up the drops and subsequently sealing the caps at chosen time intervals between 0.5 and 17 h (Nneji and Chayen, 2004). Experiments to induce and then arrest nucleation of the protein alpha crustacyanin from lobster shell were performed. Metastable conditions were found by generating a working phase diagram, screening around the conditions which gave small crystals. Trials were then set in the new crystallisation tool at metastable conditions and were allowed to evaporate for different times. The composition and volumes of drop and reservoir, the temperature and the tightness of the seal were the same in each trial. The only variable was the length of the evaporation time. The results have shown that crystal size and number is determined by the length of time that the trials have been unsealed (Nneji and Chayen, 2004). As expected, drops that were sealed throughout the experiment remained clear for several weeks. Drops that were allowed to evaporate for less than 1 h resulted in crystals appearing after 8 days while times in excess of 3 h result in showers of micro-crystals. Evaporation of 3 h gave the best results, yielding 1–8 single crystals (between 4 and 7 days) measuring at least 50 mm in each axis. Similar results were also obtained with the apoprotein crustacyanin C2 from lobster shell. This modified crystallisation tool is also useful for screening experiments, especially in the case of drops that remain persistently clear. By loosening the cap, several conditions can be explored in a single drop. This procedure has led to the detection of leads that had not been found by standard screening procedures in the case of alpha actinin actin binding domain (L. Govada, pers comm.). Dilution, as well as the induction and arrest of nucleation can also be achieved in microbatch trials. Using microbatch it is easier to automate such experiments. Saridakis et al. (1994)

Fig. 4. (a) The new design of the Nextal Crystallisation Tool containing the graduation markers. (b) A close up of an individual well.

ARTICLE IN PRESS N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

335

employed the automated microbatch technique to establish a working phase diagram for the enzyme carboxypeptidase G2. The concentrations of the protein and precipitant were varied, while pH and temperature were kept constant. The conditions for nucleation (i.e. conditions that would produce crystals if the trials were left undisturbed) were found by means of an automated system and 2 ml drops were set up as microbatch trials under these conditions. At various time intervals after set-up of the experiments, the robot was programmed to automatically insert the dispensing tip into the drops and add buffer or protein solution, thereby diluting the trials. Single diffracting crystals were routinely attained equivalent to the best, very rarely obtained without using the dilution procedure. Induction and subsequent arrest of nucleation can be performed in microbatch by setting up the trials under a very thin layer of paraffin oil, thereby allowing evaporation and hence nucleation to take place. Evaporation is later arrested by the addition of more oil to the crystallisation plate in order to seal the trials (Chayen and Saridakis, 2002). 2.4. The application of light-scattering techniques The drawback of the methods described in both microbatch and vapour diffusion is that there was no way of knowing when nucleation takes place. The time scale could only be guessed by reference to the time that it took to see the first crystals (as detailed in the dilution section). Methods that would pinpoint the appropriate time for transfer/dilution or arrest of nucleation were therefore sought. The most effective moment to intervene with a crystallisation experiment is soon after the formation of the first critical size nuclei which will eventually form the crystal. By the time nuclei or crystals can be observed under the microscope, they would have already reached a size of approximately 5 mm, by which time it is too late to act, since too many nuclei are likely to have formed (Saridakis et al., 2002). Dynamic light scattering (DLS) offers a size resolution of particles in optically transparent aqueous samples some three orders of magnitude below an optical microscope. It is therefore a useful tool for an early, non-invasive, in situ observation of a crystallisation event, before it becomes visible with a light microscope. DLS is sensitive to variations in particle size (in the range of approx. 41 nm) and interactions of protein molecules in solution (Schmitz, 1990). DLS is routinely used in many labs to assess sample mono-dispersity using dilute protein samples (D’Arcy, 1994; Ferre´-D’Amare´ and Borley, 1994; Bergfors, 1999). It has also been used successfully with lysozyme to show an increase in hydrodynamic radius as supersaturation proceeds (e.g. Mikol et al., 1990; Malkin et al., 1993; Schueler et al., 1999). Light scattering was applied for the separation of nucleation and growth by Rosenberger et al. (1993) and successfully limited the number of lysozyme crystals grown. However, the procedure involved scintillation and required 50–100 ml of sample. Simpler methods using smaller amounts of material were needed if light scattering were to be applied on a routine basis. A recent study (Saridakis et al., 2002) used DLS to monitor the crystallisation of proteins mixed with their crystallising agents in microbatch, so as to get an indication as to when to dilute the trial in order to lead it out of the nucleation phase and into the growth phase. This was achieved using a DLS-apparatus (Made by RiNA, Germany) which is able to measure a crystallisation trial as it takes place in standard cuvettes containing 10–20 ml. The study was based on the assumption that

ARTICLE IN PRESS 336

N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

changes in the aggregation profile of a supersaturated protein sample at nucleation conditions as a function of time can be used as an indicator of the induction time for nucleation. Indeed, the time at which DLS showed a significant change in the size-distribution profile of species in solution, corresponded to the time at which the solution was effectively diluted to metastable conditions, for optimal growth (Saridakis et al., 2002). In the case of vapour diffusion, the use of static light scattering has been applied to detect a change in the profile of crystallising solutions. Drops were evaporated using nitrogen gas flow and the evaporation was arrested when the light-scattering sensor detected aggregation. This approach has shown that varying the evaporation rate of the crystallisation solutions results in a smaller population of larger crystals than obtained without dynamic response (Collingthworth et al., 2000).

3. Summary and future perspectives Practical methods such as the separation of nucleation and growth of the crystallisation phases are crucial in order to obtain high-quality crystals when standard methods fail. Understandably, crystallographers were reluctant to use such methods because they were complicated, time consuming and required large quantities of material that was not readily available. This article highlights the progress that has gradually been made in the design of methods for the manipulation of the crystallisation phase diagram, reporting several methods of different complexities for separating nucleation and growth. Some of the techniques described are very simple and can be executed easily consuming less than 1 ml of sample. They can be performed without investing any extra effort or expense. Other techniques are more quantitative and sophisticated, but they require dedicated apparatus. The next step is to miniaturise and automate the procedures using light scattering which are currently performed manually or semi-automatically. Future aims include the adaptation of both dynamic and static light-scattering techniques for automated routine use in standard microbatch, vapour diffusion and other crystallisation set-ups. The post genomics era has opened up the scope for the development of new techniques and tools to overcome the bottleneck of protein crystallisation. Sophisticated instrumentation is currently being harnessed for dealing with the screening process of crystallisation. The coming years promise to bring advances in the more complicated optimisation techniques that will play a major role in raising the success rate of producing high-quality crystals.

References Ataka, M., 1993. Protein crystal growth: an approach based on phase diagram determination. Phase Transit 45, 205–219. Bergfors, T., 2003. Seeds to crystals. J. Struct. Biol. 142, 66–76. Bergfors, T.M. (Ed.), (1999). Protein Crystallization, Techniques, Strategies, and Tips. International University Line, La Jolla.

ARTICLE IN PRESS N.E. Chayen / Progress in Biophysics and Molecular Biology 88 (2005) 329–337

337

Chayen, N.E., Boggon, T.J., Cassetta, A., Deacon, A., Gleichmann, T., Habash, J., Harrop, S.J., Helliwell, J.R., Nieh, Y.P., Peterson, M.R., Raftery, J., Snell, E.H., Hadener, A., Niemann, A.C., Siddons, D.P., Stojanoff, V., Thompson, A.W., Ursby, T., Wulff, M., 1996. Trends and challenges in experimental macromolecular crystallography. Quart. Rev. Biophys. 29 (3), 227–278. Chayen, N.E., Saridakis, E., 2002. Protein crystallization for genomics: towards high-throughput optimisation techniques. Acta Crystallogr. D 58, 921–927. Chayen, N.E., Saridakis, E., El-Bahar, R., Nemirovsky, Y., 2001. Porous silicon: an effective nucleation-inducing material for protein crystallisation. J. Mol Biol. 312, 591–595. Collingsworth, P.D., Bray, T.L., Christopher, G.K., 2000. Crystal growth via computer controlled vapour diffusion. J. Cryst. Growth 219, 283–289. D’Arcy, A., 1994. Crystallizing proteins—a rational approach? Acta Crystallogr. D 50, 469–471. Ducruix, A., Giege, R. (Ed.), (1992). Crystallisation of Nucleic Acids and Proteins, A Practical Approach. Oxford University Press, Oxford. Ferre´-D’Amare´, A.R., Burley, S.K., 1994. Use of dynamic light scattering to assess crystallizability of macromolecules and macromolecular assemblies. Structure 2, 357–359. Haire Lloyd, L.F., 1996. Strategies for protein crystal growth—screening and optimization. Ph.D. Thesis,. University of London, London. Jovine, L., 2000. A simple technique to control macromolecular crystal nucleation efficiently suing a standard vapourdiffusion setup. J. Appl. Crystallogr. 33, 988–989. Korkhin, Y.M., Evdokimov, A., Shaw Stewart, P.D. (1995). Crystallisation of a protein by microseeding after establishing its phase diagram. Application Note I Douglas Instruments. Malkin, A.J., Cheung, J., McPherson, A., 1993. Crystallization of satellite tobacco mosaic virus 1. nucleation phenomena. J. Cryst. Growth. 126, 544–554. McPherson, A., 1999. Crystallization of Biological Macromolecules. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. McPherson, A., Schlichta, P., 1988. Heterogeneous and epitaxial nucleation of protein crystals on mineral surfaces. Science 239, 385–387. Mikol, V., Hirsch, E., Giege, R., 1990. J. Mol. Biol. 213, 187–195. Nield, J., Rizkallah, P., Barber, J., Chayen, N.E., 2003. The 1.45 A˚ three-dimensional Structure of C-Phycocyanin from the Thermophilic Cyanobacterium Synechococcus Elongatus. J. Struct. Biol. 141, 149–155. Nneji, A., Chayen, N.E., 2004. A crystallization plate for controlling evaporation in hanging drops. J. Appl. Crystallogr. 37, 502–503. Rosenberger, F., Howard, S.B., Sowers, J.W., Nyce, T.A., 1993. Temperature dependence of protein solubilitydetermination and application to crystallization in X-ray capillaries. J. Cryst. Growth. 129, 1–12. Saridakis, E., Chayen, N.E., 2000. Improving protein crystal quality by de coupling nucleation and growth in vapour diffusion. Protein Sci 9, 755–757. Saridakis, E., Chayen, N.E., 2003. Systematic improvement of protein crystals by determining the supersolubility curves of phase diagrams. Biophys. J. 84, 1218–1222. Saridakis, E., Dierks, K., Moreno, A., Dieckmann, M.W.M., Chayen, N.E., 2002. Separating nucleation and growth in protein crystallization using dynamic light scattering. Acta. Crystallogr. D 58, 1597–1600. Saridakis, E.E.G., Shaw Stuart, P.D., Lloyd, L.F., Blow, D.M., 1994. Phase diagram and dilution experiments in the crystallization of carboxypeptidase G2. Acta. Crystallogr. D 50, 293–297. Scheuler, J., Frank, J., Saenger, W., Georgalis, Y., 1999. Thermally induced aggregation of human transferring receptor studied by light-scattering techniques. Biophys. J. 77, 1117–1125. Schmitz, S.K., 1990. An Introduction to Dynamic Light Scattering by Macromolecules. Academic Press, New York. Stura, E.A., 1999. Seeding. In: Bergfors, T.M. (Ed.), Protein Crystallization: Techniques, Strategies, and Tips. International University Line, La Jolla, pp. 141–153. Wilson, W.W., 2003. Light scattering as a diagnostic for protein crystal growth—a practical approach. J. Struct. Biol. 142, 56–65.