Methods for separating nucleation and growth in protein crystallisation

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.     

Interaction between Plate Make and Protein in Protein Crystallisation Screening

Background

Protein crystallisation screening involves the parallel testing of large numbers of candidate conditions with the aim of identifying conditions suitable as a starting point for the production of diffraction quality crystals. Generally, condition screening is performed in 96-well plates. While previous studies have examined the effects of protein construct, protein purity, or crystallisation condition ingredients on protein crystallisation, few have examined the effect of the crystallisation plate.

Methodology/Principal Findings

We performed a statistically rigorous examination of protein crystallisation, and evaluated interactions between crystallisation success and plate row/column, different plates of same make, different plate makes and different proteins. From our analysis of protein crystallisation, we found a significant interaction between plate make and the specific protein being crystallised.

Conclusions/Significance

Protein crystal structure determination is the principal method for determining protein structure but is limited by the need to produce crystals of the protein under study. Many important proteins are difficult to crystallise, so that identification of factors that assist crystallisation could open up the structure determination of these more challenging targets. Our findings suggest that protein crystallisation success may be improved by matching a protein with its optimal plate make.     

Development of a digital video-microscopy technique to study lactose crystallisation kinetics in situ

Polarised light microscopy was employed non-invasively to monitor lactose crystallisation from non-seeded supersaturated solutions in real time. Images were continuously recorded, processed and characterised by image analysis, and the results were compared with those obtained by refractometry. Three crystallisation temperatures (10, 20 and 30 degrees C) and three different levels of initial relative supersaturation (C/C(s)=1.95; 2.34; 3.15) were investigated. Induction times using the imaging technique proved to be substantially lower than those determined using refractive index. Lactose crystals were isolated digitally to determine geometrical parameters of interest, such as perimeter, diameter, area, roundness and Feret mean, and to derive crystal growth rates. Mean growth rates obtained for single crystals were fitted to a combined mass transfer model (R(2)=0.9766). The model allowed the effects of temperature and supersaturation on crystallisation rate to be clearly identified. It also suggested that, in this set of experiments, surface integration seemed to be the rate controlling step. It is believed that a similar experimental set-up could be implemented in a real food system to characterise a particular process where crystallisation control is of interest and where traditional techniques are difficult to implement.     

Production and Characterisation of Cross-Linked Enzyme Crystals (Clecs®) for Application as Process Scale Biocatalysts

Cross-linked enzyme crystals (CLECs®) are a novel form of immobilised biocatalyst designed for application in large-scale biotransformation processes. In this work we review the production and characterisation of CLECs® prepared from three enzymes (yeast alcohol dehydrogenase I (YADHI), Candida rugosa lipase and α-chymotrypsin) over a range of crystallisation and cross-linking conditions. Optimisation and control of the crystallisation process, with respect to crystal form and enzyme activity yield, was facilitated by the use of triangular crystallisation diagrams which allowed three parameters (e.g. protein concentration, precipitant concentration and pH) to be varied simultaneously. These diagrams showed regions, or 'crystallisation windows', in which particular crystal forms or optimal activity recoveries (up to 87%) could be obtained. They also identified conditions for reproducible scale-up of the lipase crystallisation from 0.5 to 500 mL scale.

In order to evaluate the suitability of a particular batch of CLECs® for large-scale use, a hierarchy of standard tests is proposed. This is designed to expose key properties of the CLECs® relative to each other, and the free enzyme, and to minimise the number of experiments necessary to evaluate each batch of biocatalyst. In general, the CLECs® of each enzyme were found to be more resistant to harsh environmental conditions, such as extremes of temperature and pH and the presence of solvents or proteases, than the free enzymes. Cross-linking of the crystals with glutaraldehyde also yielded mechanically robust catalysts that could withstand the various forces associated with shear in agitated vessels and particle compression in repeated dead-end filtration cycles. The hierarchy of tests proposed here clearly indicated that many of the above properties were also dependent on both the crystal form and size, and the concentration of cross-linking reagent used. Accurate control of the crystallisation conditions used for CLEC® production is therefore vital as this will influence the suitability of the CLECs® for their end use.     

Lessons from crystals grown in the Advanced Protein Crystallisation Facility for conventional crystallisation applied to structural biology

The crystallographic quality of protein crystals that were grown in microgravity has been compared to that of crystals that were grown in parallel on earth gravity under otherwise identical conditions. A goal of this comparison was to assess if a more accurate 3D-structure can be derived from crystallographic analysis of the former crystals. Therefore, the properties of crystals prepared with the Advanced Protein Crystallisation Facility (APCF) on earth and in orbit during the last decade were evaluated. A statistical analysis reveals that about half of the crystals produced under microgravity had a superior X-ray diffraction limit with respect of terrestrial controls. Eleven protein structures could be determined at previously unachieved resolutions using crystals obtained in the APCF. Microgravity induced features of the most relevant structures are reported. A second goal of this study was to identify the cause of the crystal quality enhancement useful for structure determination. No correlations between the effect of microgravity and other system-dependent parameters, such as isoelectric point or crystal solvent content, were found except the reduced convection during the crystallisation process. Thus, crystal growth under diffusive regime appears to be the key parameter explaining the beneficial effect of microgravity on crystal quality. The mimicry of these effects on earth in gels or in capillary tubes is discussed and the practical consequences for structural biology highlighted.     

Lipidic cubic phase crystal structure of the photosynthetic reaction centre from Rhodobacter sphaeroides at 2.35A resolution

Well-ordered crystals of the bacterial photosynthetic reaction centre from Rhodobacter sphaeroides were grown from a lipidic cubic phase. Here, we report the type I crystal packing that results from this crystallisation medium, for which 3D crystals grow as stacked 2D crystals, and the reaction centre X-ray structure is refined to 2.35A resolution. In this crystal form, the location of the membrane bilayer could be assigned with confidence. A cardiolipin-binding site is found at the protein-protein interface within the membrane-spanning region, shedding light on the formation of crystal contacts within the membrane. A chloride-binding site was identified in the membrane-spanning region, which suggests a putative site for interaction with the light-harvesting complex I, the cytochrome bc(1) complex or PufX. Comparisons with the X-ray structures of this reaction centre deriving from detergent-based crystals are drawn, indicating that a slight compression occurs in this lipid-rich environment.     

The influence of heptane-1,2,3-triol on the size and shape of LDAO micelles. Implications for the crystallisation of membrane proteins

The presence of small amphiphiles has been found to be necessary in the crystallization of several membrane-protein/surfactant complexes. It has been suggested that the role of the small amphiphile may be to reduce the size of the surfactant belt around the protein, making the formation of crystals easier. Thus far it was not known if this would involve changes in micellar size in general or whether the small amphiphile would merely replace LDAO during crystal growth. In the present study we have used small angle neutron scattering to study mixed micelles of lauryldimethyl amine oxide (LDAO; hydrogenated and deuterated) and heptane-1,2,3-triol (HP). Our results show that with increasing overall HP concentrations mixed LDAO/HP micelles of decreasing mass and radius are formed. The composition of these micelles has been determined. HP thus may decrease the size of the surfactant belt around a protein before crystallisation by insertion into a host micelle. As HP is a 'small amphiphile' compared to the surfactants used for solubilization of membrane proteins, the curvature of the host micelle will be increased by its insertion.     

Use of RNA Tertiary Interaction Modules for the Crystallisation of the Spliceosomal snRNP Core Domain

RNA is known to perform diverse roles in the cell, often as ribonucleoprotein (RNP) particles. While the crystal structure of these RNP particles could provide crucial insights into their functions, crystallographic work on RNP complexes is often hampered by difficulties in obtaining well-diffracting crystals. The small nuclear ribonucleoprotein (snRNP) core domain, acting as an assembly nucleus for the maturation of snRNPs, plays a crucial role in the biogenesis of four of the spliceosomal snRNPs. We have succeeded in crystallising the human U4 snRNP core domain containing seven Sm proteins and a truncated U4 snRNA variant. The most critical factor in our success in the crystallisation was the introduction of various tertiary interaction modules into the RNA that could promote crystal packing without altering the core structure. Here, we describe various strategies employed in our crystallisation effort that could be applied to crystallisation of other RNP particles.     

Molecular dynamics study of polyethylene chain non-isothermal crystallisation: effects of chain length and branch structure

The molecular dynamics simulations in this work were aimed to provide a molecular insight into chain structure effects on non-isothermal crystallisation of polyethylene (PE) chains. The crystallisation behaviours were influenced by chain length and cooling rate in linear PE chain crystallisation: C100 and C150 were unable to fold into crystals. From C1000 to C3000, crystallisation abilities became stronger as chain length increased. From C5000 to C14000, chain length had no influence on crystallisation abilities. Final morphologies changed from rotator phase to single crystal domain, and to multi crystal domains as chain length increased. The formation of multi crystal domains with longer chain was easier than with the shorter chain in identical conditions. Branch content influenced not only the crystallisation kinetics but also final morphologies in non-isothermal crystallisation. The branches were defective in nucleation process, which was reflected in the crystal growth process. Crystallisation temperature, rate and crystallinity decreased, and the morphologies became disordered as branch content increased. Changes of final morphologies from single crystal domain to multi crystal domains were found under the influence of branch content and cooling rate. Trans-rich phenomenon was observed, and the trans-state population increment was prior to crystallinity increment. Crystallisation processes began at different crystallisation temperature when the trans-state populations reached a critical value which was independent of branch content.     

Target selection for structural genomics based on combining fold recognition and crystallisation prediction methods: application to the human proteome

The objective of this study is to automatically identify regions of the human proteome that are suitable for 3D structure determination by X-ray crystallography and to annotate them according to their likelihood to produce diffraction quality crystals. The results provide a powerful tool for structural genomics laboratories who wish to select human proteins based on the statistical likelihood of crystallisation success. Combining fold recognition and crystallisation prediction algorithms enables the efficient calculation of the crystallisability of the entire human proteome. This novel study estimates that there are approximately 40,000 crystallisable regions in the human proteome. Currently, only 15% of these regions (approx. 6,000 sequences) have been solved to at least 95% sequence identity. The remaining unsolved regions have been categorised into 5 crystallisation classes and an integral membrane protein (IMP) class, based on established structure prediction, crystallisation prediction and transmembrane (TM) helix prediction algorithms. Approximately 750 unsolved regions (2% of the proteome) have been identified as having a PDB fold representative (template) and an ‘optimal’ likelihood of crystallisation. At the other end of the spectrum, more than 10,500 non-IMP regions with a PDB template are classified as ‘very difficult’ to crystallise (26%) and almost 2,500 regions (6%) were predicted to contain at least 3 TM helices. The 3D-SPECS (3D Structural Proteomics Explorer with Crystallisation Scores) website contains crystallisation predictions for the entire human proteome and can be found at .     

Detergent organisation in crystals of monomeric outer membrane phospholipase A

The structure of the detergent in crystals of outer membrane phospholipase A (OMPLA) has been determined using neutron diffraction contrast variation. Large crystals were soaked in stabilising solutions, each containing a different H(2)O/D(2)O contrast. From the neutron diffraction at five contrasts, the 12 A resolution structure of the detergent micelle around the protein molecule was determined. The hydrophobic beta-barrel surfaces of the protein molecules are covered by rings of detergent. These detergent belts are fused to neighbouring detergent rings forming a continuous three-dimensional network throughout the crystal. The thickness of the detergent layer around the protein varies from 7-20 A. The enzyme's active site is positioned just outside the hydrophobic detergent zone and is thus in a proper location to catalyse the hydrolysis of phospholipids in a natural membrane. Although the dimerisation face of OMPLA is covered with detergent, the detergent density is weak near the exposed polar patch, suggesting that burying this patch in the enzyme's dimer interface may be energetically favourable. Furthermore, these results indicate a crucial role for detergent coalescence during crystal formation and contribute to the understanding of membrane protein crystallisation.     

Surface crystallisation of the plasma membrane H+-ATPase on a carbon support film for electron crystallography

Large two-dimensional crystals of H+-ATPase, a 100 kDa integral membrane protein, were grown directly onto the carbon surface of an electron microscope grid. This procedure prevented the fragmentation that is normally observed upon transfer of the crystals from the air-water interface to a continuous carbon support film. Crystals grown by this method measure approximately 5 microm across and have a thickness of approximately 240 A. They are of better quality than the monolayers previously obtained at the air-water interface, yielding structure factors to at least 8 A in-plane resolution by electron image processing. Unlike most other two-dimensional crystals of membrane proteins they do not contain a lipid bilayer, but consist of detergent-protein micelles of H+-ATPase hexamers tightly packed on a trigonal lattice. The crystals belong to the two-sided plane group p321 (a=b=165 A), containing two layers of hexamers related by an in-plane axis of 2-fold symmetry. The protein is in contact with the carbon surface through its large, hydrophilic 70 kDa cytoplasmic portion, yet due to the presence of detergent in the crystallizing buffer, the hydrophobicity of the carbon surface does not appear to affect crystal formation. Surface crystallisation may be a useful method for other proteins which form fragile two-dimensional crystals, in particular if conditions for obtaining three-dimensional crystals are known, but their quality or stability is insufficient for X-ray structure determination.     

Robotic nanolitre protein crystallisation at the MRC Laboratory of Molecular Biology

We have set up high-throughput robotic systems to screen and optimise crystallisation conditions of biological macromolecules with the aim to make difficult structural biology projects easier. The initial screening involves two robots. A Tecan Genesis liquid handler is used to transfer commercially available crystallisation reagents from 15 ml test tubes into the reservoirs of 96-well crystallisation plates. This step is fully automated and includes a carousel for intermediate plate storage, a Beckman plate sealer and a robotic arm, which transfers plates in between steps. For adding the sample, we use a second robot, a 17-tip Cartesian Technologies PixSys 4200 SynQuad liquid handler, which uses a syringe/solenoid valve combination to dispense small quantities of liquid (typically 100 nl) without touching the surface of the plate. Sixteen of the tips are used to transfer the reservoir solution to the crystallisation wells, while the 17th tip is used to dispense the protein. The screening of our standard set of 1440 conditions takes about 3 h and requires 300 microl of protein solution. Once crystallisation conditions have been found, they are optimised using a second Tecan Genesis liquid handler, which is programmed to pipette gradients from four different corner solutions into a wide range of crystallisation plate formats. For 96-well plates, the Cartesian robot can be used to add the sample. The methods described are now used almost exclusively for obtaining diffraction quality crystals in our laboratory with a throughput of several thousand plates per year. Our set-up has been copied in many institutions worldwide.     

Protein nano-crystallogenesis

We demonstrate the feasibility of growing crystals of protein in volumes as small as 1 nanoliter. Advances in the handling of very small volumes (i.e. through inkjet and other technologies) open the way towards fully automated systems. The rationale for these experiments is the desire to develop a system that speeds up the structure determination of proteins by crystallographic techniques, where most of the precious protein sample is wasted for the identification of the ideal crystallisation conditions. An additional potential benefit of crystallisation in very small volumes is the potential improvement of the crystal quality through reduced convection during crystal growth. Furthermore, in such small volumes even very highly supersaturated conditions can be stable for prolonged periods, allowing additional regions of phase-space to be prospected for elusive crystallisation conditions. A massive improvement in the efficiency of protein crystallogenesis will cause a paradigm shift in the biomolecular sciences and will have a major impact in product development in (for example) the pharmaceutical industry.     

The alpha-helical content of calmodulin is increased by solution conditions favouring protein crystallisation.

The conformation of porcine-brain calmodulin in solution has been examined by far-UV circular dichroism in the presence of 2-methyl 2,4-pentanediol, and polyethylene glycol which are used to promote the crystallisation of calmodulin. These organic compounds increase the alpha-helical content of Ca4-calmodulin to a significant degree and to a level similar to the alpha-helical content deduced from the crystal structure. These results support the view that in aqueous solution at pH 5-7, the conformation of Ca4-calmodulin is significantly different from the crystal structure and probably lacks at least a portion of the central helix. In the process of crystallisation, Ca4-calmodulin apparently adopts additional alpha-helical structure, probably due to the composition of the solution from which crystals are grown.     

Small-scale batch crystallization of proteins revisited: an underutilized way to grow large protein crystals

Growth of high-quality crystals is a major obstacle in many structural investigations. In recent years, the techniques for screening crystals have improved dramatically, whereas the methods for obtaining large crystals have progressed more slowly. This is an important issue since, although many structures can be solved from small crystals with synchrotron radiation, it is far easier to solve and refine structures when strong data is recorded from large crystals. In an effort to improve the size of crystals, a strategy for a small-scale batch method has been developed that in many cases yields far larger crystals than attainable by vapor diffusion.     

The structure and mitotic behaviour of a nuclear inclusion in a fern

Summary A study has been made of the structure and behaviour during mitosis of a crystalline inclusion within cell nuclei of roots of Dryopteris filix-max. The inclusion within the interphase nucleus is an aggregate of randomly oriented crystals. All the crystals are similar, and consist of a cubic array of particles of unit spacing approximately 100 Å. During mitosis, the inclusions are eliminated from the nucleoplasm at prometaphase. The crystals reappear within the nucleus at early interphase by a process of random crystallisation from a preformed mass of amorphous material. The results are discussed in the light of previous work on nuclear inclusions in plants and of current theories of the mode of action of microtubules.     

Expect the unexpected or caveat for drug designers: multiple structure determinations using aldose reductase crystals treated under varying soaking and co-crystallisation conditions

In structure-based drug design, accurate crystal structure determination of protein-ligand complexes is of utmost importance in order to elucidate the binding characteristics of a putative lead to a given target. It is the starting point for further design hypotheses to predict novel leads with improved properties. Often, crystal structure determination is regarded as ultimate proof for ligand binding providing detailed insight into the specific binding mode of the ligand to the protein. This widely accepted practise relies on the assumption that the crystal structure of a given protein-ligand complex is unique and independent of the protocol applied to produce the crystals. We present two examples indicating that this assumption is not generally given, even though the composition of the mother liquid for crystallisation was kept unchanged: Multiple crystal structure determinations of aldose reductase complexes obtained under varying crystallisation protocols concerning soaking and crystallisation exposure times were performed resulting in a total of 17 complete data sets and ten refined crystal structures, eight in complex with zopolrestat and two complexed with tolrestat. In the first example, a flip of a peptide bond is observed, obviously depending on the crystallisation protocol with respect to soaking and co-crystallisation conditions. This peptide flip is accompanied by a rupture of an H-bond formed to the bound ligand zopolrestat. The indicated enhanced local mobility of the complex is in agreement with the results of molecular dynamics simulations. As a second example, the aldose reductase-tolrestat complex is studied. Unexpectedly, two structures could be obtained: one with one, and a second with four inhibitor molecules bound to the protein. They are located in and near the binding pocket facilitated by crystal packing effects. Accommodation of the four ligand molecules is accompanied by pronounced shifts concerning two helices interacting with the additional ligands.     

Peptide crystal growth via sulfonate salt formation: The structure of bis-glycylglycine-1,5-naphthalenedisulfonate hydrate

Summary Continuing a line of investigations on methods for formation and growth of high-quality crystals of peptides, the glycylglycine sequence has been crystallized by evaporation methods as a salt with 1,5-naphthalenedisulfonic acid. The structure of the peptide is highly extended, and is conformationally quite similar to the structures which have been characterized for other zwitterionic and salt forms of this sequence. Thus, crystallization as a salt with this sulfonic acid has imposed no undue influence upon the molecular conformation. These results offer further indication that the preparation of peptide sulfonate salts, particularly with arene templates, may have broad general utility for crystallization of interesting sequences which until now have not been approachable in their zwitterionic forms.     

Physical properties of poly(hydroxybutyrate) and copolymers of hydroxybutyrate and hydroxyvalerate

Abstract The major physical properties of poly(hydroxybutyrate) (PHB) and its copolymers with hydroxyvalerate (PHB/HV) and those of their blends are summarised. Cases of liquid-liquid phase separation in homopolymer copolymer blends are reported and conditions where further phase separation occurs during crystallisation are documented. The crystallinity and degree of comonomer inclusion in the crystallisation of copolymers are described and quantified. It is argued that the mechanism for the inclusion of comonomer units in the crystals is of a kinetic rather than an equilibrium nature.