Hampton Research

RAMC 1997 Presentation Abstracts

Organized by Bob Cudney and Allan D'Arcy
Location: Le Bischenberg, France

T1 - Hyperthermostable Proteins: A possibility for improved crystallization?

Michael Hennig

F. Hoffmann-La Roche AG, Pharma Research PRPI-S 65/308, CH-4070 Basel, Switzerland

Hyperthermophilic organisms have adapted to life in an environment at temperatures near the boiling point of water and their proteins are able to withstand conditions that would rapidly denature proteins from mesophiles such as E. coli. In most cases, the stability is intrinsic to the protein, although some might be further stabilized in vivo by the presence of metabolites, ions, cofactors, etc. Recently, crystal structures of representatives of this class of proteins and comparison with their functionally similar proteins from mesophilic organisms have revealed determinants which could account for their different stability. Several properties such as the increased number of charged amino acids (salt bridges), increased compactness of the molecules (loop reduction) or/and firmness of the chain termini's could affect the way such proteins might crystallize. In addition, thermal stability leads to lower degradation rates and improved resistance against proteolysis, making these molecules perhaps more suitable to crystallization trials which may take weeks or months. Advantages, opportunities and drawbacks of the use of proteins from hyperthermophilic organisms for crystallographic studies will be discussed.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Hyperthermostable proteins have many charge residues on the surface and have n-terminal fixed with salt bridges.
- Hyperthermostable proteins are more compact, have reduced flexibility and have increased flexibility.
- Hyperthermostable proteins with oligomerization have improved intersubunit interactions and stabilization of chain termini. Also feature improved ligand binding.
- Hyperthermostable proteins have higher solubilities and are easier to purify.
- Hyperthermostable proteins often have improved crystal packing and reduced mosaicity.
- Active sites of thermolabile and thermostable proteins are well conserved.
- More hydrogen bonding and salt bridges result in better solubility and improved crystal contacts, and intermolecular interactions.
- Drawback: your protein may not occur in a thermophile.
Suggested Reference: Vogt & Argos, Folding & Design 2:40-46, 1997

T2 - Crystal Engineering: Rationale Design or Rolling Dice?

Glenn Dale, Martine Stihle, Brigitte D'Arcy, and Allan D'Arcy

F. Hoffmann-La Roche AG, Pharma Research PRPI-S 65/308, CH-4070 Basel, Switzerland

Protein crystallization has seen dramatic changes over the past 5-10 years in both methods and strategies. Many pharmaceutical companies have become heavily involved in protein crystallography and the requirement to produce X-ray quality crystals for knowledge based drug development has made it necessary to intensify efforts and innovation to satisfy these needs. In the past a protein crystalliser might have assumed that if enough different conditions were tested that any given protein might crystallize, despite the availability of many and varied crystallization screens this assumption can no longer be made. Light scattering studies for example have shown that certain proteins would have little chance of crystallizing due to the presence of large molecular weight, non specific aggregates. In our lab we have used a number of methods for modifying proteins such as complex formation, limited proteolysis or enzymatic de-glycosylation to improve the chances of obtaining usable crystals. We are now beginning to use the tools of molecular biology for "crystal engineering" in order to produce modified proteins that maintain specific activity but are better suited to crystallographic studies. Improving solubility or introducing improved crystal contacts are two important goals in these studies. We have initiated a program of site directed mutagenesis of surface residues on a DNA binding protein which has proved difficult to crystallize. The various mutants which could be isolated in a soluble form were screened for crystallization characteristics using an automated microbatch diffusion method. Initial results and some conclusions will be presented.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Random approach: Make random surface mutations to introduce more charged residues. One then selects soluble constructs with activity then screens for crystallization conditions. In general every mutant had an effect on crystallization and only a limited number of mutants were required to get good crystals.
- Rational Approach: Introduce charged residues to the surface of the protein. For example, Lysine to Arginine. Replace hydrophobic residues with alanine. If a structure is available, review structure to guess which mutations might improve crystal contacts.

T3 - HIV Integrase--Improving Crystal Quality Through Molecular Biology Technics

Anne M. Hassell and Lisa Shewchuk

Glaxo Wellcome Inc. 5 Moore Drive, Research Triangle Park, N.C. 27709, USA

HIV integrase catalyzes the insertion of viral DNA into the host genome through a distinct series of DNA cutting and joining reactions. In the 3' processing step, two nucleotides are removed from each 3' end of the double-stranded, blunt-ended DNA produced from reverse transcription. The recessed 3' ends of the viral DNA are then covalently joined to the 5' ends of the target DNA, completing the integration process and permanently infecting the host cell. The fact that HIV integrase has no known counterpart in human cells makes it a very attractive target for the design of anti-HIV agents. However, structure-based drug design efforts were hindered by the insolubility of the enzyme. Dyda et al (Science 266, 1981-1986 [1994]) overcame the solubility difficulties encountered in their crystallization efforts by substituting LYS for PHE185. Although this mutation resulted in a three-dimensional structure of HIV integrase core domain (50-212), the active site loop of the structure was disordered. We will describe how changing the expression plasmid and the NH2-terminal histidine tag improved the crystal quality, enabling us to obtain a 3-dimensional structure with an ordered active site loop.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Working with HIV integrase. Problem: Active loop disordered and cacodylate buffer is bound to the protein. Did later find that cacodylate altered protein structure. This is rare and has not stopped them from using cacodylate as a crystallization buffer.
- Made two changes.
--- Changed to pRSETA plasmid for expression
--- Left His tag intact on protein
- Result?
--- More efficient purification
--- Obtained crystals in more conditions, some without cacodylate
--- Crystals with improved diffraction, ordered active loop
- Tip: Put BioRad Chelex 20 in fraction collection tubes to scavenge nickel in order to prevent protein precipitation.

T4 - Light scattering: a crystal ball for crystallization or just another way to waste a few mg of protein?

Terese Bergfors

Uppsala University Dept of Molecular Biology, BMC Box 590, 751 24 Uppsala, Sweden

Native gel electrophoresis and solution light scattering are two methods for detection of protein heterogeneity resulting from oligomerization. These two techniques were used to characterize 35 proteins to determine the correlation between oligomeric homogeneity and crystallization. This study suggests confirms some of the trends described in an early paper based on l7 proteins1 but contains a larger sample size and additional comparison with native gels.

85% of the proteins characterized as monodisperse by light scattering produced crystals, although not always suitable for data collection. Interestingly, proteins containing bimodal dispersions performed just as badly as multiply polydisperse proteins in terms of crystallization. This suggests that it is the fact of polydispersity, not its degree, which is detrimental. Nevertheless, polydispersity was not an ipso facto deterrent to crystallization as some highly polydisperse proteins did still produce crystals. These exceptions will be discussed as well as the cases where monodisperse proteins failed to crystallize.

Finally, a comparison between native gels and light scattering measurements showed that oligomeric homogeneity was a more important factor in crystallization than purity from other protein contaminants. References: 1. Zulauf, M. and D'Arcy, A. (1992) J. Crys. Growth 122, 102-106.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Compared dynamic light scattering (DLS) to native gels to assay protein for crystallizability.
- DLS is a better predictor (81%) of crystallizability than native gel (61%).
- When monomodal, a protein has an 81% probability of crystallizing.
- When bimodal or polydisperse, crystallization is possible but there is a much lower probability of crystallization.
- When polydisperse, one should proceed with screening, but not waste too much time or material. If no crystals are obtained with a bimodal or polydisperse protein one should try to clean up the protein or change the protein.
- It is possible to have a monodisperse protein and obtain no crystals. When this occurs try deglycosylating the protein, proteolytically modifying the protein, screen additives or try other things that will manipulate sample-sample and sample-solvent interactions.
- If one has a polydisperse protein, one can try changing the temperature, the pH, and salt in an attempt to obtain a monodisperse protein.

T5 - Just So Stories on Gal6 and Other Complexes (With apologies to R. Kipling)

Julie Rosenbaum1, Troy Messick1, Stephen Albert Johnston2 and Leemor Joshua-Tor1

1Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, NY 11724 USA,
2UT-Southwestern Medical Center, Dallas TX 75235-8573 USA

Resistance to antineoplastic drugs, whether intrinsic or acquired, is a central problem in treatment of human cancers. Bleomycin hydrolase is a cysteine protease discovered by its ability to hydrolyze the anticancer drug bleomycin and thus to limit its use in cancer therapy. The recent discovery of homologs of this intracellular, large hexameric protein in yeast and bacteria implies a conserved cellular function for this protein. In humans, this protease is expressed in all tissues tested and at elevated levels in several tumors. In mice, expression is highest for newborn mice and decreases with age. The crystal structure of the yeast bleomycin hydrolase, Gal6, which we have determined has revealed several unique features for these proteases. These include a nucleic-acid binding activity and an unusual positioning of the completely conserved C-terminus of the protein in its own active site cleft. The full length 300 kD yeast homohexameric protease was expressed using an overexpression system in yeast which will be described as well as using a histidine tag in E. coli. The techniques for obtaining crystals suitable for data collection will be described along with the Dynamic Light Scattering results used to monitor preparations of additional protein complexes.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Tip: Had a drop with a small, poorly diffracting crystal. Added a small amount of water to the drop and resealed. The crystal partially dissolved and then grew back larger with improved diffraction qualities.
- Has experienced better success with a TEV His tag site then an EK site.

T6 - Control of nucleation by a simple temperature shift method

Empirical screening methods are often successful in producing crystals. Subsequent control of nucleation to give a plentiful supply of X-ray quality crystals may be problematic and more difficult to achieve than obtaining crystals in the first instance. "Hit and miss" procedures such as seeding or screening for a magic additive are often employed in the search for optimal conditions. A more systematic approach may be used by establishing the phase diagram for the protein under study. This may be done easily and rapidly using the automated microbatch technique. Although crystal solubility remains difficult to measure, the supersolubility curve (above which spontaneous nucleation occurs) can be observed easily and gives the essential information needed to select crystallization conditions. The nucleation and growth processes may be separated by two methods: by diluting from initial conditions of high supersaturation to lower supersaturation for growth after a chosen incubation time; or, as described here, by incubating a solution at a temperature where it is supersaturated and then changing to an appropriate temperature for growth. The temperature shift strategy devised for this study was more convenient and practically easier to perform than the batch dilution method as no additional dispensing and mixing steps were involved. In any crystallization where there is significant variation in protein solubility with temperature, this strategy should be considered if obtaining large single crystals is a problem.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
-If protein solubility varies significantly with with temperature, try:
--- Nucleating crystals at one temperature (the temperature where the sample is least soluble), then...
--- Growing crystals at a second temperature where sample has increased solubility.
--- Benefit: No additional dispensing or mixing required. Found microbatch method was most convenient. No condensation with temperature change.

T7 - Drop-Drop: Macromolecular crystallization using Micro-Volume Vapor Diffusion

George T. DeTitta and Joseph R. Luft

Hauptman-Woodward Medical Research Institute, 73 High Street, Buffalo, New York 14203-1196, USA

Consider a typical macromolecular vapor diffusion crystallization experiment. A protein solution (evaporator) of approximately 10 microliter volume is suspended in a closed system over a reservoir solution (condenser) of approximately 1000 microliter volume. The condenser volume is ³ infinitely² larger that the evaporator and controls the endpoint concentration of the evaporator. To change the endpoint concentration in the evaporator one must change the chemical conditions of the reservoir solution. We present a modification of the traditional vapor diffusion experiment called ³drop-drop². The experiment is set up using organosilane treated 8 x 30mm Tite Seal vials (Kimble Glass, Inc., #D-830) with polyethylene plug caps. A condenser solution of up to 100 microliter volume is placed in the cap, while the evaporator solution is placed in the bottom of the treated glass vial. Variation in evaporator endpoints can be achieved by varying the condenser and evaporator solution volumes. Using a 1 M NaCl solution as the evaporator and a 2 M NaCl solution as the condenser varying volumes for one or both solutions, the endpoint can be varied in a smooth progression from 1 to 2 M NaCl. Gravimetric analysis has shown the vials to have only a 4% loss of water after a period of 19 months. The power of this technique is evident when using the Crystal Screen (Hampton Research) solutions. The sampling of endpoints can help delineate amorphous from microcrystalline precipitate. Endpoints of conditions producing ³microcrystalline precipitate², can be screened by setting up a drop-drop experiment and varying only the condenser drop volume. Drop-drop crystallization experiments have been conducted using toth Hen Egg White Lysozyme and Beef Liver Catalase. In addition, the results of kinetics experiments using 1 M NaCl as the evaporator and 2 M NaCl as the condenser show the rate of equilibration to be slower than similar experiments conducted in a Linbro plate. In summary, the method represents (1) significant cost savings when using expensive reservoir reagents, (2) greater efficiency in endpoint sampling, varying condenser volume rather than chemical composition, (3) slower equilibration rates which can be beneficial for crystallization. Work supported in part by NASA grant NAG8-1152.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Method uses a small Kimble Titeseal Vial (Fisher Scientific)
--- Use a 1 to 100 microlier reservoir volume
--- Why? Saves time & reagent. Allows optimization of kinetics of equilibrium by a blind screen approach. How? By changing the volume of the reservoir, one can change the endpoint of the drop. Method allows for slower drop/reservoir equilibration than traditional Linbro/VDX plate with 1 milliliter reservoir volume.

T8 - Effect of a magnetic field on the protein crystallization

Tohoku University, Institute For Materials Research Katahira 2-1-1, Aoba-ku, Sendai 980-77, Japan

Abstract: Crystallization of hen egg-white lysozyme and horse spleen ferritin was carried out under steady and uniform magnetic field of 10 T and compared with the crystals grown under 0 T. It is clearly demonstrated that a magnetic field reduced the number of the nuclei and not only oriented the crystals but also modified the habit of the protein crystals. The present experimental result indicates that application of a magnetic field is an efficient method for growing a small number of large crystals.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Applied a 0 to 10 tesla magnetic field to lysozyme for three days at 20°C.
- Found that:
--- Magnetic field changed the crystal habit
--- Magnetic field changed crystal orientation Why? Perhaps diamagnetic anisotropy of crystal
--- 0 tesla field resulted in 160 crystals, where 10 tesla field resulted in 80 crystals.
--- Cost of magnet about $100,000

T9 - The Role of Oil in Macromolecular Crystallization

Imperial College of Science, Technology & Medicine, London SW7 2BZ, United Kingdom

The crystal growth of proteins is a complicated process which is dependent on numerous factors. This talk highlights the unique contribution of oil as a major parameter in protein crystallization. The utilization of oil for protein crystallization was originally initiated in order to enable the dispensing and incubation of very small crystallization samples using the microbatch method in which crystals are grown in 1 - 2 microliter drops of a mixture of a protein and crystallizing agents. The primary role of the oil was to act as an inert sealant to prevent evaporation of the small-volume trials. Experimental evidence has revealed that the oil itself can play an important part in the outcome of a crystallization experiment by affecting the crystallization process throughout its stages (of nucleation, growth and the stability of the resulting crystals). A wide range of experiments which exploit the presence of oil to aid protein crystal growth are presented. This talk focuses on protein crystals, although the methods described also apply to other biological macromolecules.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Described microbatch, oil layer and containerless methods for crystal growth
- Methods utilize paraffin and/or silicon/fluorinated oils
- Advantages: Control of nucleation, reproducibility, protection of sample, makes temperature studies easier.
- Disadvantages: Shock nucleation, does not work with all organics (no thymol or dioxane), harvesting crystals can be tricky.
- Methods are OK with detergents and membrane proteins

T10 - Crystallization of the membrane protein cyclooxygenase-2: Beauty is only skin deep

Stefania Di Marco, John P. Priestle and Markus G. Grutter

Novartis Pharma AG., CH-4002 Basel, Switzerland

A recombinant apoprotein form of the membrane enzyme cyclooxygenase-2 (COX-2), has been purified and reconstituted with heme. The holoprotein has been crystallized in complex with the selective inhibitor 6-(2,4-difluorophenoxy)-5-methylsulfonylamino-1-indanone. Brownish-yellow rod-shaped crystals of dimensions 0.8 x 0.1 x 0.1 mm3 were obtained after one week using polyethylene glycol of various molecular weights, with the hanging drop method. Brownish-yellow rhombohedral-shaped crystals of dimensions around 0.1 x 0.15 x 0.05 mm3 were obtained after one week to two months with 2-methyl-2,4-pentanediol (MPD) using the hanging drop method. Crystals from the PEG experiments were difficult to handle and were easily destroyed during mounting. Those crystals that survived mounting seldom diffracted beyond 9 angstroms. Crystals from MPD were easier to handle but also diffracted to only around 8 angstroms resolution. Efforts to increase the size of both crystal forms resulted in lengths of up to 1 mm, but no corresponding increase in width and breadth and therefore no increase in diffraction power. Finally, an improved crystal form was obtained using the sitting drop method of vapor diffusion with PEG 2000 monomethyl ether as precipitant, in the presence of the nonionic detergent beta-octylglucoside. Despite the poor morphology of these crystals, they diffracted to about 2.5 angstroms resolution. The crystals are orthorhombic, belonging to the space group P21212 with cell dimensions a = 209.56, b = 71.28 and c = 93.82, and diffract to 2.5 angstrom resolution. The asymmetric unit (Vm = 2.65 3 Da-1, solvent volume = 54%) suggests the presence of two COX-2 monomers, in agreement with the dimeric structure of the detergent-solubilized protein found by light scattering and size exclusion chromatography. The molecular replacement solution confirmed that the asymmetric unit contains one dimer of COX-2.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Described crystal with good morphology yet no diffraction. Described the same protein in a crystal with poor morphology that gave good diffraction. Moral: Check all the crystals, even the ugly ones.
- Check to see if protein is active n detergent prior to using the detergent for crystallization studies.
- Remember and consider micelle size when selecting a membrane size cut-off for sample concentration. Might also concentrated detergent!
- Use temperature change, salt type/concentration, and precipitant type/concentration to prevent/manipulate phase separation in drops.
- Try PEG MME and low molecular weight PEGs such a PEG 200-4000

T11 - Lipidic Cubic Phases: A Novel Concept for the Crystallization of Membrane Proteins

Ehud M. Landau, Gabriele Rummel, Eva Pebay-Peyroula and Jurg P. Rosenbusch

Biozentrum, University of Basel, Klingelbergstr. 70, CH-4056 Basel, Switzerland and Institut de Biologie Structurale/ Universite Joseph Fourier, 41 av. des martyrs, F-38027 Grenoble, France

A molecular level understanding of the mechanisms of action of membrane proteins requires elucidation of their structures to high resolution. To date, only a few high resolution structures of membrane proteins have been solved, reflecting the major stumbling block in this endeavor - the routine production of well-ordered three-dimensional crystals. We have devised a novel concept for the crystallization of membrane proteins by exploiting the properties of bicontinuous lipidic cubic phases.1 This membrane-mimetic system, consisting of lipid, water and protein in appropriate proportions, forms a transparent 3-D curved bilayer matrix, which is pervaded by two non-connected aqueous channels. Membrane proteins, once inserted into this complex array, diffuse laterally to nucleate, and eventually to yield well-ordered crystals. Bacteriorhodopsin crystals were obtained from bicontinuous cubic phases, but not from micellar systems, implying a critical role of the continuity of the diffusion space (the bilayer) on crystal growth. Hexagonal bacteriorhodopsin crystals diffract to high resolution, with a space group P63, and unit cell dimensions of a=b=61.76A, c=104.16A, a=b=90o and g=120o, and one monomer per asymmetric unit. References: 1. E. M. Landau and J. P. Rosenbusch, Proc. Natl. Acad. Sci. USA, 93, 14532 (1996) 2. E. Pebay-Peyroula, G. Rummel, J. P. Rosenbusch and E. M. Landau, submitted

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Is a membrane mimetic system composed of lipid, water and protein
- Allows for continuity in diffusion space and the system is pervaded by non-connected solvent channels.
- Properties: Membrane like material, structured, highly viscous, facilitate diffusion, table, transparent - System used to obtain crystals of bacteriorhodopsin; structure solved
- Small crystals, 50 by 5 microns

T12 - Intelligent Computational Aids for Crystal Growth

John M. Rosenberg1, Patricia A. Wilkosz1, K. Chandrasekhar1, Devika Subramanian2, Daniel Hennessy3 and Bruce Buchanan3

1Depts. of Biological Sciences & Crystallography, University of Pittsburgh USA
2Dept. of Computer Science, Rice University USA
3Intelligent System Laboratory, University of Pittsburgh USA

During the course of crystallization experiments, substantial data accumulate on the conditions that lead to unsuccessful, partially successful and successful crystallizations. The project described here seeks to provide computational tools for the collection and interpretation of that data. Specifically, the goals of this project are to design, implement, and test and intelligent, interactive, electronic assistant for crystallographers that will facilitate the trial and error process of growing diffraction quality crystals of biological macromolecules based on: 1. Archiving of crystallization experiments. 2. Accessing the database¹s crystallization trials including generation of new experimental protocols. 3. Inducing empirical theories that capture regularities in the data. 4. Suggesting plausible next steps in a series of crystallization trials. The initial ³front end² will be demonstrated. We also invite volunteers to test the software and to contribute data for the next stage of the project, which is the application of artificial intelligence methods. Our initial analytical efforts utilized the data in the BMCD database. We found that significant improvements in the statistical interpretation of the BMCD required classifying macromolecules according to a analyses, including the Student T-test, to the BMCD data. As one representative example, we asked whether the distribution of macromolecular concentration was systematically different for the protein subclasses. We found that the heme containing proteins and the membrane proteins stand out as significantly different from the rest of the protein families. Additional data will be reported. How can statistical results like this be incorporated into the design of crystallization experiments? We have developed software that calculated Bayesian probabilities for any combination of crystallization parameters, using data retrieval from the archive, currently the BMCD. The calculated probabilities are used to bias the selection of data from an incomplete factorial design such that the more probable combinations are sampled more densely that the less probable ones. This feature has been included in the software to be demonstrated and made available, as described above.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- PC based software for the design, execution, recording and analysis of protein crystallization experiments.
- Features a screen design module module which utilizes BMCD data and internal crystallization data to design biased factorial/sparse matrix screens.

T13 - The Cyberlab C-200: A Fully Automated System for Protein Crystallization Vapour Diffusion Experiments

Tom Friedlander

Cyberlab, Inc., 36 Del Mar Drive, Brookfield, CT 06804, USA

The past 10 years have seen dramatic changes in the field of protein crystallization. Recombinant DNA technology has made it possible to produce most soluble proteins in large quantities and commercially available sparse matrix type screens have been routinely and successfully used to establish initial crystallization conditions. In addition many more pharmaceutical research laboratories are becoming involved in X-ray crystallography in the area of lead identification and optimization, where rapid identification of crystallization conditions is of great importance. The need to automate some of the more tedious and time consuming tasks involved in obtaining protein crystals was recognized a number of years ago with the introduction of the first "crystallization robots" of varying levels of sophistication ( K.Ward, B. Reubin, N. Jones, ICN, Douglas Instruments and others). We have developed a system which is designed to mimic the manual set up of a vapour diffusion experiment with a minimum of human intervention, including greasing of Linbro plate wells and inversion of the coverslip for the hanging drop technique. The system can be programmed to run up to 6 Linbro type plates without supervision and using protein volumes as low as 1ul. The development and a description of the work stations hardware and software will be presented.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Robotic workstation which automatically designs, dispenses and mixes crystallization reagents.
- Greases 24 well plates (Linbro and VDX)
- Pipets drops, flips cover slides
- Hanging and sitting drop

T14 - Shrinking protein crystals - a way to higher resolution

Patrick Cramer

European Molecular Biology Laboratory (EMBL) Grenoble Outstation c/o ILL, B.P. 156 F-38042 Grenoble, France

Biological macromolecules often crystallize in weakly diffracting forms which are unsuitable for structure determination. Recently, several cases have been reported where such crystals could be transformed resulting in diffraction to higher resolution. The extension of the diffraction limit involved dehydration and a contraction of the unit cell. A co-crystal form of the eukaryotic transcription factor NF-kappaB P52 homodimer:DNA complex showed anisotropic diffraction which limited the resolution of the obtained data to 3.5 A. Soaking of these crystals in harvest buffer solutions containing ytterbium ions and increasing amounts of PEG 400 caused shrinking which could be followed under the microscope. The packing rearrangements involve shortening of the b-axis by 10 A (8%). The transformation extended the diffraction limit to better than 2.0 A resolution.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- 3.4 angstrom crystal was exposed to PEG 400 and Yb(NO3)3 resulting in 2.0 angstrom crystal. How? Possible tightening of the crystal packing?
- A literature review showed that in such experiments there is typically a decrease in the solvent content of the crystal by 3 to 25%. There is a change in the cell volume, decreasing by 3 to 14%. In general, an increase in diffraction resolution.
- Other protocols for shrinking crystals include letting the drop dry out (duh!) dehydrate the crystal with calcium chloride, potassium chromate, or high molecular weight PEG, and adding heavy metals to the drop.

T15 - Flash freezing isomorphous xenon or krypton derivatives of protein crystals

O. Sauer1, R. Dutzler2, C. Kratky

1 Department of Structural Biology, Institute of Physical Chemistry, University of Graz, Austria 2Department of Structural Biology, Biocenter of the University of Basel, Switzerland

Higher electron density, reasonable anomalous dispersion and relatively good solubility of Xenon and Krypton suggest to use these gases in protein crystallography. If ³specific² binding sites exist in proteins, Xe and Kr are used as heavy atoms for MIR, SIRAS, and MAD. If the concentration of these gases in the solvent regions is sufficiently they alter the solvent contrast. In the case of a non uniform dielectric distribution in the mother liquor, one expects an accumulation of the rare gas in apolar regions with a low dielectric constant. Such regions are present in crystals of membrane proteins due to the detergents used to solubilize the protein. Hence the electron density of these regions are enhanced specifically and the contrast against the protein and the remaining solvent part is altered. Therefore, Xenon and Krypton could be used as probes for imaging hydrophobic regions in crystals of membrane proteins. Xenon and Krypton bind to proteins due to very weak dispersion forces. To observe a rare gas protein interaction, a sufficiently high concentration of the gas in the mother liquor surrounding protein in the crystal must be provided. This can be achieved by exposing the protein crystal to high pressures of the desired gas. Depending on the structure of the protein, on the compactness of the crystal and especially on the composition of the mother liquor, the necessary pressure to attain usable isomorphous derivatives ranges from a few bar toa hundred bar or more. Due to the smaller polarizability of Kr as compared to Xe, the former has to be applied with at least 30% higher pressures to obtain the same occupancy. For contrast variation studies, the crystal must be exposed to very high pressures (typically 50 to 100 bar). To keep the established binding condition during data collection on a diffractometer, either the pressure must be kept elevated or the crystal must be flash-frozen, before desorption of the dissolved gas from the mother liquor occurs. Since the former method has serious disadvantages (particularly for diffraction experiments with Cu-K alpha radiation), we concentrated on the latter and developed a pressure device which allows one to pressurize protein crystals up to 110 bar of rare gas. The device is constructed in such a way that a release of the pressure within one or tewo seconds is possible. Immediately afterwards the crystal is shock frozen in liquid nitrogen without a recognizable loss of dissolved gad. Compared to pressure cells which maintain elevated gas pressure during the X-ray experiment (where the crystal is sealed in X-ray quartz capillaries), the method circumvents strong absorption effects of compressed gas surrounding the crystal, the possible formation of clathrates, radiation damage and permits much higher gas pressures, which are not limited by the mechanical stability of the quartz capillaries. The feasibility of producing isomorphous Xenon derivatives with this method is shown with crystals of sperm whale metmyoglobin. The feasibility of enhancing the electron density of specific regions in a crystal is demonstrated with low resolution difference maps of OmpF-porin crystals from E. coli following exposure to Xe or Kr pressures of 60 bar.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- A device for the creation and freezing of xenon/krypton derivatives was demonstrated. Performed at 50 to 100 bar pressure.
- Xenon and krypton derivatives can be used to image hydrophobic regions
- Generally does not interfere with crystal contacts

T1 - Polymers as nucleants under high salt conditions

Alexander McPherson

University of California, Irvine
Department of Molecular Biology and Biochemistry, Irvine, California 92697 U.S.A.

In the course of crystallizing a variety of antibodies and Fab fragments, it has been found that a useful precipitant for growing crystals is lithium sulfate at concentrations of one to two molar. Nucleation in the salt seldom occurs, however, unless some polymer such as PEG 4000 is also present at low concentration, 0.1 % to 0.5 %. With the inclusion of polymers, nucleation is reproducible and growth proceeds normally. From a preliminary investigation of about 25 different potential polymeric nucleants in lithium sulfate solutions, using a selected set of immunoglobulins and Fabs, it appeared that Jeffamine® ED - 2000 Reagent was most effective. In any case, it was observed that effective enhancement of nucleation only occurred when the polymer was at sufficiently low concentration that no little or no phase separation of polymer from salt solution was evident. A more extensive exploration of conditions is currently underway and results of these additional experiments will be reported.

Jeffamine® is a registered trademark of the Huntsman Petrochemical Corporation

T16 - Hardware for cryocrystallography

Elspeth Garman and Mike Pickford

Laboratory of Molecular Biophysics, Oxford University, United Kingdom

Many laboratories have developed their own hardware and `gizmos' to facilitate cryocrystallography data collection, as well as for crystal transfer, storage and retrieval. The equipment which has been developed at the Laboratory of Molecular Biophysics over the last few years will be displayed together with a poster on their use and design features. This hardware includes: an alignment tip for a cryostat nozzle, pins for loops, `top-hats' for loops, magnetic goniometer adaptors, a loop making jig, a removable arc for a goniometer and various tools which we have found useful.

Bob's Unabashed Notes:
Warning: These notes were taken under the influence of the beautiful French countryside, food and drink. While these notes are perhaps 1% accurate we recommend you defer to the authors for the whole truth and nothing but the truth. That said, enjoy the info!
- Cryo? Data collection at 100°K. Why? Better and more data, all data from a single crystal, MAD is easier.
- If using a PEG < 4000 simply increase the PEG concentration for cryo or try adding low molecular weight PEGs.
- Most cases can use 15-255 glycerol
- If using salt as the precipitant try MPD or ethylene glycol
- Make sure you do not dilute the mother liquor when adding cryoprotectant. For example, to get a 30% glycerol cryoprotectant concentration in a drop add 0.5 parts of 2x reservoir, 0.3 parts 100% glycerol and 0.2 parts water.
- In general rapid transfer to cryoprotectant is better. But osmotic shock is high with quick transfers. Double edged sword problem.
- One can try sequential soaks. But rather than move the crystal, aspirate the mother liquor, replace with cryoprotectant mixture, repeat. Handling the crystal can sometimes increase mosaic spread.
- Size of crystal is important. Surface/volume is the key. If surface/volume is greater than 12 try cryo with a smaller crystal.
- Competitive with glycerol can happen. If it does try changing from glycerol to MPD.
- For the record, Elspeth was inhumanely tortured by the slider projector and electrical cords during the presentation but managed to survive. Sorry Elspeth! For the complete and most excellent story on cryo by Elspeth see: E.F. Garman and T.R. Schneider, 1997, J. Appl. Cryst. 30, 211-237.