SP.01 What we Have Learned from Structures of the Ribosome Venki Ramakrishnan MRC, Cambridge, United Kingdom Ever since its discovery in the 1950s, the ribosome has been the object of study by labs world-wide, because of its central role in the translation of genetic information into proteins. However, its large size meant a long delay in the determination of its structure, despite the fact that the crystals of the ribosome were first obtained around 1980. I shall discuss our determination of the structure of the 30S subunit, and more recent work on high-resolution functional complexes of the entire ribosome, as well as insights into ribosome function from these structures. In particular, the structural basis of decoding, in which the tRNA corresponding to a codon on mRNA is accurately selected, will be discussed.

AW.01 Crystallography - is the Gold Standard Getting Tarnished? David Watkin University of Oxford, Oxford, United Kingdom X-rays were discovered by Roentgen in 1895, diffraction of X-rays by von Laue in 1912, and the crystal structure of sodium chloride was elucidated by Bragg in 1913. Initially structures were determined by trail and error methods, but in 1936 Patterson published his interpretation 2 of the F synthesis - the Patterson function. After that, there was no looking back, and the determination of molecular structures using X-ray diffraction developed rapidly. By the end of the 1960's most of the theoretical background which we now use had been worked out. The technique had been almost completely explored in just 30 years, so that many of the senior practitioners had experienced all the stages in its development. Because they had done much of the work by hand, from estimating intensities through to laborious calculations, they appreciated the significance of all the different kinds of data they were working with. They understood the risks in taking short cuts, and were able to make informed decisions about the advantages and disadvantages in alternative procedures. Because these people really understood what they were doing, their results (within the limits of the instrumentation available) were soundly based and secure - they were The Gold Standard of analytical techniques.

I trained as a student in the 1960's and so had a reasonably formal education in crystallography. I experienced the transition from cameras to analogue diffractometers, to serial diffractometers, TV and image plate diffractometers, through to ccd and now pixel array detectors. Computers have reduced calculation times by perhaps four orders of magnitude (one could now do in one year what would have taken 10,000 years). Most of the theories developed by the pioneers are routinely available in brilliant programs. While crystallography in the broad sense still has many exciting frontiers, much molecular structure analysis has become routine, and some would say of a deteriorating quality. The IUCr, in trying to maintain standards, has the unenviable tasks of trying to distinguish good work on poor samples from poor work on good samples.

02.01.1 How Low Can You Go: Investigation of the Effects of Redundancy on Absorption Correction Michael Takase Massachusetts Institute of Technology, Cambridge, MA, United States Semi-empirical absorption correction using the multi-scan approach [1] is based on the comparison of equivalent reflections. It can be expected that this method would work best if the number of equivalent reflections in a given dataset is high. Similarly, semi-empirical absorption correction should not work very well for datasets with low or very low redundancies. In an attempt to determine the critical value of minimum redundancy (if, in fact, there is such a value) below which effective semi-empirical absorption correction is no longer possible, we are analyzing a number of datasets from a variety of different samples. Variables to be examined are crystal size, presence or absence of heavy atoms, Laue symmetry and wavelength of X-radiation

[1] R.H. Blessing, Acta Crystallogr., Sect A 1995, 51, 33-38.

02.01.2 How are we doing? A Review of Small Molecule Crystallography based upon Data Mined from the Cambridge Structural Database. Joseph Reibenspies Texas A & M University, College Station, Texas, United States It is human nature for one to stand back from time to time and ask themselves: how am I doing? The question is of course very subjective; however it is fair and reasonable to ask such a question of a group as a whole. To measure progress and thus provide an answer to the title question one can examine a representative instrument, such as the Cambridge 1 Structural Database . The database consists of crystallographic information from 1923 to the present and encompasses organic and inorganic molecular compounds. Information such as 2 structure counts, residual factors, volumes, percent errors, total disorder and bond distance precision provides statistical data that can be used to measure progress in terms of quantity, quality and complexity. With this information in hand one can also be so bold as to predict the direction the science of crystallography is taking and what the future holds. At the end of the discussion we may ask ourselves How are we doing? and the answer may come as a bit of a surprise.

1

Allen, F. R. (2002). Acta Cryst. B58, 380388.

2

Flippen-Anderson, J. L., J. R. Deschamps, Gilarid, R.D., George, C. (2001) Crystal Engineering 4, 131-139.

OQMOPMR r ? \ ` ?k h ??n\ ?b K?t ??y K?y K?r \ \ Y?\ \ K? \K??

The validation of crystal structures via the checkCIF software began in late 1997. Before then, it was up to the practitioner to check carefully that all was well with each structure determination; a manageable task when only a few dozen or less structures were determined per year and a wise professional crystallographer was at hand for guidance. Nowadays, diffractometers, in some cases in do-it-yourself labs, have such high throughput that mere mortals are overwhelmed by the amount of data being accumulated. To cope with the number of structures being determined, intelligent tools to automate the routine parts of the experiment are welcome and new automatic structure determination software is appearing. Can we now sit back, push a button and enjoy seeing a finished (routine) structure appear before our eyes together with a clean validation report, then assume all is okay? There may be aspects in any structure determination that the validation tools cannot evaluate or give appropriate feedback about before it is too late. It is still necessary to keep a watchful eye on the entire experiment and to think carefully and critically about the results. Every crystal has its own peculiarities, some of which may need special treatment, so even at the data collection stage one should be watchful of what is unfolding and be prepared to take appropriate action.

A recent addition to the checkCIF suite is the ability to validate the structure factor listing against the corresponding CIF. This enables users to confirm that their archived structure factor file corresponds with the refinement run used to generate the CIF. The tests may also give feedback about, among other things, overlooked twinning and other inconsistencies within the CIF that might arise, for example, from incorrect editing of an existing CIF after a new refinement. Access to the service is at: http://journals.iucr.org/services/cif/checking/checkcifhkl.html

02.01.4 High Pressure Cryocooling at MacCHESS Chae Un Kim , Irina Kriksunov , William A. Miller , Mike Cook , Doletha M. E. Szebenyi , Sol 2,3 M. Gruner1 1 1 1 1 1

MacCHESS (Macromolecular diffraction at CHESS), Ithaca, NY, United States, Cornell High 3 Energy Synchrotron Source, Ithaca, NY, United States, Physics Department, Cornell University, Ithaca, NY, United States A novel high-pressure cryocooling technique for macromolecular crystallography has been developed and explored at the Macromolecular Diffraction Facility at the Cornell High Energy Synchrotron Source (MacCHESS) [1]. The method involves cooling macromolecular crystals to cryogenic temperatures (~ 100 K) in high-pressure (up to 200 MPa) helium gas. Applications include successful cryocooling with little or no penetrating cryoprotectant. The method has been extended to Kr/Xe single-wavelength anomalous dispersion (SAD) phasing, native sulfur SAD phasing, and preparation of cryocooled crystal samples in capillaries. A mechanism involving high-density amorphous (HDA) ice is used to explain why the method works. Surprising results include visualization of ligands which could not be seen using other methods [2, 3], and insight into the phases of water in a protein crystal [4]. The high pressure cryocooling method is available to researchers with suitable crystals: See [http://www.macchess.cornell.edu/MacCHESS/about_macchess.html#Pressur e]. References [1] C. U. Kim, R. Kapfer, and S. M. Gruner (2005), Acta Cryst. D61, 881-890. [2] R. A. Albright, J. -L. V. Ibar, C. U. Kim, S. M. Gruner, and J. H. Morais-Cabral (2006), Cell 126, 1147-1159. [3] J. F. Domsic, B.S. Avvaru, C. U. Kim, S. M. Gruner, M. Agbandje-McKenna, D. N. Silverman, and R. McKenna (2008), J. Biol. Chem. 283, 30766-30771.

2

[4] C. U. Kim, B. Barstow, M. W. Tate, and S. M. Gruner (2009), Proc. Natl. Acad. Sci., 106, 4596-4600.

02.01.5 Progress in Using Short Wavelength Radiation for Chemical Crystallography Juergen Graf , Bernd Hasse , Francesca Fabbiani , Thomas Schulz , Dietmar Stalke , Holger 3 1 Ott , Carsten Michaelsen1 3 1 1 2 2 2

Incoatec GmbH, Geesthacht, Germany, University of Goettingen, Goettingen, Germany, Bruker AXS GmbH, Karlsruhe, Germany

2

Combining synthetic multilayer mirrors with microfocus X-ray sources (rotating or stationary target) has become a standard with in-house X-ray sources for single crystal diffraction as well as a number of applications in powder diffraction. The maximum angle of incidence at which a multilayer mirror reflects is significantly smaller for higher energy radiation, such as Mo-K or Ag-K radiation than it is for Cu-K radiation. This is why synthetic multilayer mirrors traditionally have been used for Cu-K radiation or softer wavelengths. Modern deposition technology, however, allows for the reproducible production of high quality multilayer mirrors with smaller d-spacing. In consequence these mirrors reflect higher energy radiation at larger angles of incidence. Combined with the latest generation of microfocus sealed tubes this provides new high-performance low-power X-ray sources for shorter wavelengths.

We will present selected results on the use of these low-power consumption, highperformance sources in small molecule and high-pressure crystallography.

02.01.6 A Hybrid Pixel Detector in the Home Laboratory: Prospects for Better Data Joseph Ferrara, Colin Acheson, Angela Criswell, Pierre Le Magueres, James Pflugrath, Katsunari Sasaki Rigaku Americas Corp, The Woodlands, TX, United States We have begun using a hybrid pixel detector (HPD), specifically the Dectris Pilatus 100K, in home lab single crystal X-ray diffraction experiments. In order to assess the utility of such a device for the home lab, we have studied the performance of this device for both small molecule and protein data collection experiments with copper radiation. We will present results comparing HPD data collection to conventional CCD data collection as well as results comparing conventional data collection to shutterless data collection in terms of data quality and increased throughput.

07.01.1 First Results of Femtosecond Protein Nanocrystallography

Mark Hunter , Petra Fromme , Rick Kirian , Uwe Weierstall , Bruce Doak , Henry Chapman , 1 John Spence Arizona State University, Tempe, AZ, United States, Hamburg, Germany1 2

1

1

1

1

1

2

CFEL/University of Hamburg,

Serial crystallography has been used to show the first proof-of-principle for femtosecond nanocrystallography at the Linac Coherent Light Source, a 2keV pulsed X-ray laser at SLAC -15 which provided 3-300 femtosecond (10 s) pulses. The intensity of the X-ray pulses exceeds rd 3 generation sources by 12-orders of magnitude, yet the pulses are so short that X-ray diffraction data are collected before the sample is destroyed. In serial crystallography, X-ray diffraction is collected from a stream of fully hydrated protein nanocrystals in their mother liquor. The jet introduced nanocrystals of Photosystem I, a complex membrane protein with a mass of 1056000 Da consisting of 36 protein subunits and 381 cofactors, to the femtosecond X-ray beam produced at the LCLS. Individual diffraction patterns, read out at 30 Hz, could then be indexed and assembled into a working data set. Over six million diffraction patterns from Photosystem I nanocrystals were collected, and diffraction was recorded to the detectorlimited resolution of 9. The experiments indicate that in this diffract-and-destroy mode, even a 70 fs pulse may terminate before detectable radiation damage or spot fading occur, to the available resolution. The impact and potential of the LCLS for future structural determinations of membrane proteins will be discussed. This project is a large international collaboration, involving the CAMP group from three Max Plank Institutes and ASU physics. PIs include H. Chapman, P. Fromme, I. Schlichting, B. Doak, U. Weierstall, J. Uhlich, A. Barty, L. Struder, D. Rolles, the LCLS staff and the ASG team.

07.01.2 Crystal structure of the membrane fusion protein CusB from Escherichia coli Edward Yu Iowa State Universiry, Ames, United States Gram-negative bacteria, such as Escherichia coli, frequently utilize tripartite efflux complexes belonging to the resistance-nodulation-division family to expel diverse toxic compounds from the cell. These systems contain a periplasmic membrane fusion protein that is critical for substrate transport. We here present the x-ray structures of the CusB membrane fusion protein from the copper/silver efflux system of E. coli. This is the first structure of any membrane fusion proteins associated with heavy-metal efflux transporters. CusB bridges the inner membrane efflux pump CusA and outer membrane channel CusC to mediate resistance + + to Cu and Ag ions. Two distinct structures of the elongated molecules of CusB were found in the asymmetric unit of a single crystal, which suggests the flexible nature of this protein. Each protomer of CusB can be divided into four different domains, whereby the first three domains are mostly -strands and the last domain adopts an entirely helical architecture. Unlike other known structures of membrane fusion proteins, the -helical domain of CusB is folded into a three-helix bundle. This three-helix bundle presumably interacts with the periplasmic domain of CusC. The N and C-termini of CusB form the first -strand domain, which is found to interact with the periplasmic domain of the CusA efflux pump. Atomic + + details of how this efflux protein binds Cu and Ag were revealed by the crystals of the CusBCu(I) and CusB-Ag(I) complexes. The structures indicate that CusB consists of multiple binding sites for these metal ions. These findings reveal novel structural features of a membrane fusion protein in the resistance-nodulation-division efflux system, and provide evidence that this protein specifically interacts with transported substrates.

07.01.3 Crystal structure of the ectodomain complex of the CGRP receptor, a Class-B GPCR, reveals the site of drug antagonism Ernst ter Haar, Christopher Koth, Norzehan Abdul-Manan, Lora Swenson, Joyce Coll, Judith Lippke, Christopher Lepre, Miguel Garcia-Guzman, Jonathan Moore Vertex Pharmaceuticals Incorporated, Cambridge, MA, United States The calcitonin gene-related peptide (CGRP) is a potent vasodilator directly implicated in the pathogenesis of migraine. Its receptor (CGRP-R) is a heterodimer containing the calcitonin receptor-like receptor (CLR), a class B GPCR, and RAMP1, a receptor activity-modifying protein. We have solved the crystal structure of the CLR/RAMP1 N-terminal ectodomain heterodimer, revealing how RAMPs bind to and modulate the activities of the CLR GPCR subfamily. We have also determined the structures of CLR/RAMP1 in complex with antagonists olcegepant (BIBN4096BS) and telcagepant (MK0974). Both drugs act by blocking access to the CGRP binding cleft at the interface of CLR and RAMP1. These structures reveal how small molecules bind to and modulate the activity of a class B GPCR, and highlight the challenges of designing potent receptor antagonists for the treatment of migraine and other class B GPCR-related diseases.

07.01.4 Resolving the Structures of Membrane Pores Formed by Antimicrobial Peptides

Huey Huang Rice University, Houston, Texas, United States Antimicrobial peptides (AMPs) are ubiquitous components of the innate immune systems found in all plants and animals. Soon after their discovery in the 80s, they were found to kill microbes by forming pores in the microbial membranes. Since the conventional antibiotics have been facing the serious issue of resistance, this new type of antimicrobials has attracted a great deal of interest. However their molecular mechanisms as well as the structures of their pores have been controversial. In this talk I will describe how we used neutron scattering and X-ray diffraction to resolve the structural and mechanism issues. In particular we have developed a new MAD procedure to resolve the phase problem of diffraction.

07.01.5 Interaction of Lipid Monolayers and Single Supported Bilayers with Cholera Toxin: X-ray and Neutron Reflectometry and Grazing Incidence X-Ray Diffraction Studies Jaroslaw Majewski , Tonya Kuhl , Chad Miller , Erik Watkins1 1 2 2,1 1,2 2

Los Alamos National Laboratory, Los Alamos, NM, United States, University of California Davis, Davis, CA, United States Biological membranes are critical components of functioning cells and many bacterial toxins bind to and gain entrance to target cells through specific interactions with membrane components. Using surface sensitive neutron and x-ray reflectometry and grazing-incidence diffraction we were able to follow the process of cholera toxin attack on a model lipid mono- and bi-layer. In-plane and out-of-plane changes in 2-D packing of cholera toxin molecules and the lipid membrane were investigated. We followed the process of the toxins assault on the monolayer in time. A firm understanding of the molecular mechanisms by which cholera toxin penetrates and translocates across a membrane will stimulate the design of possible interventional therapies to prevent infections that use the same mechanism to enter the cell. Furthermore, a similar mechanism could be employed to transfect cells with a desired therapy.

07.01.6 Lipid membrane-mediated 2D assembly of proteins and viruses at liquid interfaces Masafumi Fukuto , Suntao Wang , Sumit Kewalramani , Matthew Lohr , Zhongwei Niu , 2 2 1 Giang Nguyen , Qian Wang , Lin Yang Brookhaven National Laboratory, Upton, NY, United States, University of South Carolina, Columbia, SC, United States Biomolecular nanoparticles (BNPs), such as proteins and viruses, are ideal nanoscale building blocks because of the intrinsic monodispersity in their size, shape, and surface properties. In particular, BNPs bound to a lipid monolayer at a solution-vapor or solution-substrate interface are well suited for investigating ordered 2D assembly of nanoscale objects. Using in situ grazing-incidence x-ray scattering, we have recently studied density-driven 2D crystallization of BNPs for two types of BNPmembrane interactions, one based on specific ligand binding and the other based on electrostatic interactions. For the first system, the 2D assembly of the protein streptavidin (SA) on a biotinbearing lipid monolayer was studied as a function of the surface density of biotin, a protein-binding ligand. The results of detailed x-ray scattering and optical Brewster-angle microscopy measurements reveal that the 2D crystallization of the lipid-bound SA occurs as a density-driven first-order phase transition. Significantly, the threshold biotin density for inducing the 2D crystallization is found to be roughly equal to the density of the ligand-binding sites in the SA crystal. Moreover, the extracted protein adsorption isotherm indicates that the fully bound state of SA, corresponding to two biotinlipids per protein, is achieved already below the threshold biotin density. These results demonstrate that in addition to a well-defined molecular orientation, high lateral packing density is essential to the 2D crystallization of proteins. For the second system, the electrostatic 2D assembly of cowpea mosaic virus (CPMV) on a mixed cationic-zwitterionic (DMTAP/DMPC) lipid monolayer was studied as a function of the subphase pH and the membrane charge density. GISAXS data show that 2D crystals of CPMV are formed above a threshold membrane charge density and only in a narrow pH range just above CPMV's isoelectric point, where the charge on CPMV is expected to be weakly negative. The particle density for the 2D crystals is similar to that for the densest lattice plane in the 3D crystals of CPMV. The results demonstrate that the 2D crystallization is achieved in the part of the phase space where the electrostatic interactions are expected to maximize the adsorption of CPMV onto the lipid membrane.1 2 1 1 1 1 2

01.01.1 LOUIS DELBAERE and the Early Days of Crystallography in Edmonton, Alberta. Title: Michael James University of Alberta, Edmonton, Alberta, Canada Louis Delbaere made many seminal contributions to the advancement of Canadian Crystallography during his days in Edmonton. He solved several structures of monosaccharides while he was a post doctoral fellow in Ray Lemieuxs laboratory. He also made major contributions to the understanding of the structures of the blood-group determinants. Louis also made major contributions to protein crystallography and to solving the first protein structure done in my laboratory in Edmonton. He was one of the main motivating forces in solving the crystal structures of the A and B peptidases from Streptomyces griseus. He also collaborated with Gary Brayer on the structure of the alphalytic protease. He worked with I-Nan Hsu and Theo Hofmann on the structure of the first aspartic peptidase to be solved, that of Penicillopepsin.

01.01.2 Louis Delbaere: Friend and Colleague J. Wilson Quail Department of Chemistry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada For years to come, many Canadians who never knew Louis will benefit from his efforts. He left a mark on many people and organizations.

01.01.3 Louis Delbaere and the International Union of Crystallography Sine Larsen International Union of Crystallography, Chester, United Kingdom, University of Copenhagen, Copenhagen, Denmark When Louis too early passed away in 2009 he had served as a member of the IUCr Executive Committee (EC) for 14 months. Both his election to the EC and the selection of Montreal as the venue for the 2014 IUCr congress demonstrated the respect and appreciation that the international crystallographic community held for Louis. Having known Louis from the time we both started our crystallographic career it was a personal pleasure to work with him in the EC the short time we had together. This presentation will focus on the role Louis played in the international crystallographic society and the impact of his work for the IUCr.1 2

01.01.5 Surface Constrained Protein Nanotubes for Bionanoelectronics Gerald Audette, Stephanie Lombardo, Agnesa Shala York University, Toronto, ON, Canada Powering the next generation of implantable devices will rely systems that are more biologically accessible. To achieve this, we must harness the power of proteins and interface them with scaffolds, creating unique bio-nanosystems. Bionanoelectronics are interface a protein system to a conducting surface, thereby connecting them to micro fuel sources such as biofuel cells. One system we are currently investigating is the type IV pilus from Pseudomonas aeruginosa, a nanofibre composed of multiple copies of a single protein subunit, the type IV pilin. In the presence of a hydrophobic surface or solution, engineered pilin monomers oligomerize into soluble, high molecular weight structures protein nanotubes (PNTs). P. aeruginosa pilins, pili and pilin-derived PNTs have been shown to bind both biotic surfaces (cells) as well as abiotic surfaces such as stainless steel. We have recently examined the oligomerization of PNTs onto alkylthiol functionalized gold surfaces. PNTs oligomerized from surface constrained alkylthiols have been observed to be several micrometers in length with an average diameter of 36 3 nm. In comparison with reported values for the diameter of native type IV pili and PNTs in solution (~6 nm), the average observed diameter of surface oligomerized PNTs suggests a multiple PNT clustered filament on the gold surface. Current research targets the directed engineering of the pilin monomer to facilitate metal ion binding for bionanowire development and patterned PNT oligomerization on gold surfaces via differential functionalization of the gold surface with a variety of alkylthiols.

01.01.6

AlgK and AlgE form the outer membrane secretin of a novel bacterial expolysaccharide secretion system Lynne Howell , John Whitney , Carrie Keiski , Maria Amaya , Mirela Neculai , Patrick 1 1 1 1 1 4 Yip , Laura Riley , Joel Weadge , Francis Wolfram , Yura Lobsanov , Howard Robinson , Lori 3 5 Burrows , Dennis Ohman The Hospital for Sick Children, Toronto, ON, Canada, University of Toronto, Toronto, ON, 3 4 Canada, McMaster University, Hamilton, ON, Canada, Brookhaven National Laboratory, 5 Upton, NY, United States, Virginia Commonwealth University Medical Center, Richmond, VA, United States Pseudomonas aeruginosa is the predominant pathogen associated with chronic lung infections in Cystic Fibrosis patients. During colonization of the lung, P. aeruginosa converts to a mucoid phenotype characterized by the overproduction of the exo-polysaccharide alginate. Ten proteins, encoded by the algD operon, have been implicated in alginate biosynthesis. Since polymerization requires protein components in both the inner and outer membrane, these proteins are believed to form a large multi-protein complex that spans the cell envelope and facilitates export of the polymer. To understand how alginate is exported we have undertaken structure-function studies on two outer membrane components of this system, the lipoprotein, AlgK, and the -barrel porin, AlgE, both of which are essential for the production of high molecular weight alginate. Complementation and subcellular fractionation studies have shown that AlgE is required for alginate secretion, and that AlgK plays an important role in localizing AlgE to the outer membrane; a finding that suggests that the two proteins may interact directly. Structures of AlgK and AlgE have been determined at 2.5 and 2.3 resolution, respectively, and reveal that AlgK is an all -helical solenoid protein with at least 9.5 tetratricopeptide-like (TPR) protein-protein interaction motifs. This topology suggests that AlgK may play a role in assembly of the alginate biosynthetic complex. AlgE is a monomeric 18-stranded -barrel that shows remarkable structural similarity to members of the substrate-specific OprD-family of outer membrane channels. The occlusion of AlgE s electropositive conduit by extracellular loop L2 and periplasmic loop PL8, suggests that, unlike in other channels, conformational changes may be required to facilitate alginate export. Bioinformatic analyses have also revealed that the proteins BcsA, PgaA and PelB, involved in the production and export of cellulose, poly- -1,6-N-Acetyl-D-glucosamine and Pel exopolysaccharide, respectively, share the same topology as AlgK/E. Together our data provide a model for alginate export and suggest that AlgK/E represent the periplasmic and outer membrane components of a new type of outer membrane secretin that differs from canonical bacterial capsular polysaccharide secretion systems.1 2 1,2 1,2 1,2 1 1

01.01.7 Dynamics of Siderophore Reception at the Staphylococcus aureus Cell Surface Jason Grigg , John Cooper , Johnson Cheung , David Heinrichs , Michael Murphy1 2 1 2 2 2 1

University of British Columbia, Vancouver, BC, Canada, University of Western Ontario, London, ON, Canada Staphylococcus aureus is a devastating bacterial pathogen. It thrives in the human body by scavenging iron with secreted small molecule chelators called siderophores. S. aureus synthesizes two polycarboxylate-type siderophores, staphyloferrin A (SA) and staphyloferrin B (SB). Siderophore-iron complexes are bound at the cell surface by lipoprotein receptors (HtsA and SirA) and shuttled to membrane permease for import. We have determined the crystal structures of both receptors in their apo and siderophore-bound complexes. The SA receptor, HtsA, binds a single SA-Fe complex with nM affinity. Structures have been determined of apo-HtsA (1.35 resolution), HtsA-SA-Fe in an open conformation (1.30 resolution) and closed conformation (2.2 resolution). The closed ternary complex structure was obtained from a twinned crystal (P21 with operator l,-k,h). Together these structures reveal that instead of more typical domain movement, significant local conformational changes isolated to three loops in the C-terminal domain coincide with SA-Fe binding. A ~12 C shift of Tyr239 brings it into H-bonding distance of SA, enclosing the SA-Fe complex in the pocket. The conformational change also results in a ~ 2.8 shift of Glu250, located at the apex of the domain where it likely forms a salt-bridge with the permease, allowing discrimination between apo and holo receptors without large scale domain movement. Importantly, the structure of SA-Fe is also well defined in the binding pocket allowing assignment of chiral centers, one of which differs from previous predictions based on model compound properties. The apo-SirA (2.2 resolution) and SirA-SB-Fe ternary complex (1.7 resolution) structures have been determined, again allowing definition of the SB-Fe structure. SirA belongs to the same structural family as HtsA and also undergoes localized conformational changes following ligand binding. However, the large structural change occurs on the opposite side of the binding pocket to HtsA, highlighted by a ~13 shift of Asp263 C to enclose SB-Fe. Localized conformational changes bring two arginine residues in H-bonding distance of SB as well as a reorientation of Glu245 by ~1.6 , which analogous HtsA is predicted to allow discrimination of ligand bound and apo-SirA.

07.02.1 Ultrafast synchrotron and laboratory in situ powder diffraction studies on the synthesis of functional materials John Evans Durham University, Durham, United Kingdom Traditionally the vast majority of functional inorganic materials have been synthesised under conditions of thermodynamic control - the "heat and beat" method of solid state synthesis. Such methods are, of course, unsuitable for the preparation of metastable materials. In this presentation I'll discuss how in-situ powder diffraction methods have been crucial in allowing the preparation of several metastable inorganic framework materials, and in suggesting the topotactic mechanisms that allow their isolation. Examples will include the preparation of new molybdate and phosphate materials followed using laboratory based techniques.[1]

I'll also discuss new ultrafast powder diffraction experiments performed at beamline ID11 at the ESRF in which full quantitative Rietveld refinement is possible on reacting ceramic systems with a time resolution down to 0.1 seconds. These experiments have allowed us to identify conditions under which we can prepare the negative thermal expansion material ZrMo2O8 directly from its constituent oxides. The ability to rapidly scan temperature and composition space on a reaction that is complete in time periods of ~10 seconds has allowed us to discover a small temperature window at around 1500 K in which this material is thermodynamically stable, and to optimise conditions for its synthesis.

I'll also discuss data analysis methodologies which allow one to extract the maximum information from large bodies of diffraction data collected during in-situ experiments.

[1] Lister, S.E., Evans, J.S.O., unpublished results. [2] Readman, J.E., S.E. Lister, L. Peters, J. Wright, and J.S.O. Evans, Direct Synthesis of Cubic ZrMo2O8 Followed by Ultrafast In Situ Powder Diffraction. Journal of the American Chemical Society, 2009. 131(48): p. 17560-17562.

07.02.2 Structural behaviour of (Mg,Fe )(AlSi)O3 perovskite at pressures of the Earth s lower mantle Tiziana Boffa Ballaran , Alexander Kurnosov , Konstantin Glazyrin , Marco Merlini , Daniel J. 1 Frost Bayerisches Geoinstitut, Bayreuth, Germany, Dept. Earth Science, University of Milano, Milano, Italy Throughout the bulk of the Earths lower mantle, MgSiO3-perovskite is expected to contain significant proportions of both Al and Fe, with Fe potentially in both 2 and 3+ oxidation states. Several studies have reported that ferrous iron exists in an intermediate spin state throughout the Earths lower mantle whereas ferric Fe may be involved in a high-spin low-spin transition. Changes in spin states in silicate perovskite can be expected to affect the coordination polyhedra around the iron atoms and influence therefore the elastic properties of this mineral. In order to quantify the effect of such spin transitions on the structural behaviour of perovskite, we have synthesised at the Bayerisches Geoinstitut single-crystals of a very Fe,Al-rich perovskite using a multi-anvil technique. The crystals are of an excellent quality as indicated by their sharp diffraction profiles and lack of twinning. High-pressure single-crystal X-ray diffraction has been performed at the beam line ID09 at the European Synchrotron Radiation 3+ Facility (ESRF). A single crystal of (Mg,Fe )(AlSi)O3 (longest dimension 25 microns) has been loaded in a diamond anvil cell with ruby as pressure standard and He as pressure transmitting medium. Intensity data have been collected at different pressures up to 75 GPa and isotropic structural refinements always converged with discrepancy factors smaller than 4%. Octahedral and dodecahedral sites have similar compressibility and the orthorhombic distortion only sligthly increases with pressure. However, the octahedral tilting along the c axis shows a change in behaviour above 50 GPa.1 2 1 1 1 2 3+

07.02.3 Probing the high-pressure behaviour of H2SO4 and MgSO4 hydrates with neutrons A. Dominic Fortes , Ian G. Wood , Matthew G. Tucker1 1,2 1,2 3 2

Department of Earth Sciences, University College London, London, United Kingdom, Centre 3 for Planetary Sciences at UCL/Birkbeck, London, United Kingdom, ISIS Facility, Rutherford Appleton Laboratory, Chilton, Oxfordshire, United Kingdom

Hydrates of sulfuric acid, and of magnesium sulfate, have been reported on the surfaces of the Galilean satellites of Jupiter [1], and models suggest that these hydrates will be abundant in their deep interiors [2], consequently experiencing modest hydrostatic pressures. Investigation of their high-pressure behaviour is important since there are likely to be changes in both the hydrogen bond network as well as possible changes in ion speciation, and pressure-induced dehydration. Such phase changes may influence heat transport inside icy planetary bodies, and hence control their overall structure and evolution. Using neutron powder diffraction, we have carried out studies upon a range of hydrates relevant to the internal structure and dynamics of icy satellites, including deuterated isotopologues of sulfuric acid 8-, 6-, and 4-hydrate, and magnesium sulfate 11-, and 7-hydrate, at pressures up to 4 GPa in the Paris-Edinburgh press. Much of our earlier work is summarised in reference [3]. In sulfuric acid tetrahydrate, we have identified two monoclinic high-pressure polymorphs, SAT-II and SAT-III, in addition to the low-pressure tetragonal phase. SAT-II is formed by warming SAT-I above 235 K at 550 MPa, and this has been successfully recovered to atmospheric pressure at 50 K. SAT-III has been observed over the range 1.63.9 GPa, melting at 380 K at 3.9 GPa. However, sulfuric acid tetrahydrate is extremely difficult to crystallise at high-pressure, requiring deep undercooling. In the MgSO4-hydrates we have now observed a reproducible sequence of phase transitions, in agreement with ultrasonic wave-velocity observations reported elsewhere [4], in which pressure-induced dehydration occurs. At 295 K, MgSO4 7D2O (synthetic epsomite) undergoes its first transition at 1.2 GPa to a lower hydrate + aqueous solution, this phase being stable over only a narrow pressure range (0.2 GPa), before the onset of the next phase transition and the growth of ice VII. Similarly, MgSO4 11D2O (synthetic meridianiite), when compressed at 240 K, breaks down to a lower hydrate + ice VI at ~ 0.9 GPa. We report the status of our work to interpret the high-pressure behaviour of these materials, including recent attempts to recover the products of various phase transitions to ambient pressure for better characterisation. References [1] Orlando et al. (2005) Icarus 177, 528-533: McCord et al. (2001) Science 292, 1523-1525. [2] Kargel (1991) Icarus 94, 369-390. [3] Fortes & Choukroun (2010) Space Sci. Rev. 10.1007/s11214-010-9633-3. [4] Gromnitskaya et al. (2010) High Press. Res. 30, 51-54.

07.02.4 Crystallographic studies of medical gases adsorbed in metal-organic frameworks Russell Morris University of St Andrews, St Andrews, United Kingdom Metal-organic frameworks comprise one of the most exciting classes of solids in current science. These highly porous solids have been of particular interest as gas sorbent materials. Many of these studies have concentrated on adsorbing hydrogen, carbon dioxide, methane and other gases of interest for energy and environmental applications. However, metalorganic frameworks are also exceptional materials for the adsorption, storage and delivery of medically important gases, such as nitric oxide (NO). In this presentation I will describe our crystallographic studies of NO adsorbed into several different framework solids, and how this information helps us to understand the particular properties of the materials in question. To complete these studies we primarily use single crystal X-ray diffraction studies at synchrotron sources (Daresbury, UK and the ALS, USA). The experiments are completed using a specially designed environmental gas cell that allows the single crystals to be thermally activated and then loaded with gas while it is on the diffractometer.

References. 1. Xiao B et al. Chemically blockable transformation and ultraselective low-pressure gas adsorption in a non-porous metal organic framework Nature Chemistry 289-294 (2009). 2. McKinlay AC et al. Exceptional behavior over the whole adsorption-storage-delivery cycle for NO in porous metal organic frameworks J. Am. Chem. Soc. 130, 10440-10444(2008).

3. Warren JE et al A prototype environmental gas cell for in situ small-molecule X-ray diffraction J. Appl. Crystallogr. 42, 457-460 (2009).

07.02.6 High pressure behavior of a family of 2+ 2+ 2+ 2+ [DMA][M(CHO2)3], M = Mg , Co , Mn and Zn1 2

perovskite-related3

metal1,4

formates,

Lauren A. Borkowski , Hyunsoo Park , Stephen Moggach , John B. Parise1 2

Mineral Physics Institute, Stony Brook University, Stony Brook, NY, United States, 3 Department of Chemistry, Indiana University, Bloomington, IN, United States, School of 4 Chemistry, The University of Edinburgh, Edinburgh, Scotland, United Kingdom, Departments of Chemistry and Geosciences, Stony Brook University, Stony Brook, NY, United States2+

The M formates crystallize in a perovskite-related structure, ABX3, where the A cation is dimethylamine (DMA), and some display multi-ferroic behavior at low temperatures. The inorganic perovskites display remarkable structural flexibility, often dominated by polyhedral distortion and tilting. We are interested in exploring the high pressure behavior of this new class of materials and comparing this to known perovskite compression mechanisms. Previous work at low temperature shows [DMA][Mg(CHO2)3] transforms from R-3c to C2/c. 2+ 2+ Single crystal and powder high pressure experiments on the Mg and Co analogs show that the high pressure behavior of these materials consists of tilting of the metal octahedra as well as a partial ordering of the DMA molecules on the A site.

07.02.5 The XIPHOS diffraction facility for experiments at ultra low temperatures1 1

extreme1

sample

environments:1

In-house2

Craig Robertson , Michael Probert , Jonathan Coome , Andres Goeta , Brian Michell , Judith 1 Howard1

Durham University, Durham, United Kingdom, Bruker AXS Inc., Madison, WI, United States

2

The XIPHOS (X-ray Interface for Photo-Induced High pressure lOw temperature Structural studies) diffraction facility has been developed to collect diffraction data within a range of 1 sample environments. The system couples a Bruker direct drive Mo rotating anode generator operating at 5.4 kW with the latest Helios focusing optics. This source is mounted on a four circle Huber goniometer equipped with an APEX II CCD detector. To reach ultra rd low temperatures, an APD 202E Displex cryogenic refrigerator with an additional 3 JouleThompson stage has been installed, allowing for temperatures as low as 2 K to be maintained. This new diffraction facility will be presented in detail; furthermore, low temperature calibration, crystal centring and the results of recent experiments will be presented.

References: 1) Probert, M. R.; Robertson, C. M.; Coome, J. A.; Howard, J. A. K.; Michell, B. C.; Goeta, A. E. Submitted, J. Appl. Cryst., February 2010 (hx5107).

07.02.7 'The Big Squeeze on Porous Materials.' Stephen Moggach The University of Edinburgh, Edinburgh, United Kingdom Recent interest in gas storage materials has led to a plethora of papers on the synthesis of [1, 2] metal organic framework materials (MOFs). Structural variation in MOFs can be achieved through chemical modification, with accompanying changes in pore size and shape (and therefore internal surface area) giving rise to an increasingly diverse array of sorption properties. Such sorption measurements are performed at pressures up to 0.01GPa, though what effect higher pressures have on the framework is relatively unknown. A sub-family of MOFs are the so called zeolitic imidazolate framework (ZIF) materials. ZIFs, related to zeolites through the 145 angle subtended at the bridging imidazolate ligand, are of increasing interest. Their tuneable pore size, chemical robustness and thermal stability combine the most desirable features of conventional MOF and zeolite structures, making them ideal candidates for gas storage applications. Over the last 10 years, developments in highpressure single-crystal diffraction techniques have allowed us to study much larger [3] compounds than was previously possible. The scope for pressure to change material properties has been demonstrated in previous work on amino acids and molecular magnets. This work focussed on tuning hydrogen bonding interactions or magnetism respectively. Here, we present the effect of pressure on porous molecular materials, in particular ZIF-8 (Zn(MeIM)2, MeIM = 2-methylimidazolate). On increasing pressure, we were not only able to [4] tune the pore size, but also the pore content of this material through pressure modification.

[1]R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M. OKeeffe, O. M. Yaghi, Science 2008, 319, 939. [2]R. Kitaura, G. Onoyama, H. Sakamoto, R. Matsuda, S.-i. Noro, S. Kitagawa, Angew. Chem., Int. Ed. 2004, 43, 2684. [3]S. A. Moggach, D. R. Allan, S. Parsons, J. E. Warren, J. Appl. Crystallogr. 2008, 41, 249. [4]S. A. Moggach, T. D. Bennett, A. K. Cheetham, Angew. Chem., Int. Ed. 2009, 48, 7087.

07.03.1 Data Collection, Reduction and Semi-automatic Structure Solution with HKL-3000 Wladek Minor , M. Cymborowski , M. Chruszcz , Z. Otwinowski , D. Borek1 2 1 1 1 2 1

Univ. of Virginia, Charlottesville, VA, United States, UT Southwestern Medical Center, Dallas, TX, United States HKL-3000 integrates data collection, data reduction, phasing, and model building to significantly accelerate the process of structure determination, and on average, minimize the number of data sets and crystals required for structure solution. Execution of the package merges several modules and software applications into the structure determination pipeline. There are modules for experimental control of some beamlines and home instruments, data reduction, phasing by SAD/MAD or molecular replacement, fast model building, and initial refinement. The system is being developed and tested in the high-throughput environment of the Midwest Center for Structural Genomics (MCSG) and Center for Structural Genomics of Infectious Diseases (CSGID). The robustness of HKL-3000 has improved considerably over time and currently over 1000 structures have been determined with it. The continuous advancement of the decision-making procedures within HKL-3000 have made it the system of choice for MCSG and CSGID projects. Transforming raw images into a solved structure (with 70% of the model built) in 10-15 minutes is no longer a surprise, but a routine operation for crystals that diffract to 2.5 or better. Our experience with the determination of hundreds of structures by experimental phasing methods helped us to establish rules for the best approaches when the available data fall into three categories: unsolvable with current data, borderline and easy. Current work concentrates on improving the approach to borderline cases of structure determination rather than optimizing intermediate calculations for easy cases, thus shifting borderline cases into the easy category and unsolvable into borderline. An important implication is that simple experimental protocols are sufficient in most cases and may even be optimal for the most challenging ones. Feedback from fast preliminary structure solution proved to be one of the critical components of success. References Minor W, Cymborowski M, Otwinowski Z, Chruszcz M (2006) Acta Crystallographica Section D: Biological Crystallography62:859-66. Kirillova O, Chruszcz M, Shumilin IA, Skarina T, Gorodichtchenskaia E, Cymborowski M, Savchenko A, Edwards A, Minor W (2007) Acta Crystallographica Section D: Biological Crystallography63:348-54. Otwinowski Z, Borek D, Majewski W, Minor W (2003) Acta Crystallographica. Section A: Foundations of Crystallography59:228-34. Zheng H, Chruszcz M, Lasota P, Lebioda L, Minor W (2008) Journal of Inorganic Biochemistry102(9):1765-76. Chruszcz M, Wlodawer A, Minor W (2008) Biophysical Journal95(1):1-9. Wlodawer A, Minor W, Dauter Z, Jaskolski M (2008) Febs Journal275:1-21. Otwinowski Z, Minor W (1997) Methods in Enzymology275:307-326.

07.03.2 CMDDENZO and CMDXDS: Single-Line-Command-Driven Automated Data Processing at SER-CAT1,2 1,2 3

User1,2

Interfaces1,2

for

Zheng-Qing Fu , Zhongmin Jin , Andy Howard , John Chrzas , Jim Fait , Unmesh 1,2 1,2 1,2 1,2 1,2 Chinte , John Gonczy , Rod Salazar , John Rose , Bi-Cheng Wang Department of Biochemistry & Molecular Biology, University of Georgia, Athens, United 2 3 States, SER-CAT, APS, Argonne National Lab, Argonne, United States, BCPS, Illinois Institute of Technology, Chicago, United States As part of ongoing SER-CAT efforts to monitor data quality on-the-fly we have developed command-line-driven user interfaces (UI's) CMDDENZO and CMDXDS, which exploit functions in some of the widely-used data reduction packages such as 1) 2) 3) 4) 5) DENZO/SCALEPACK , D*TREK , SPGR4D, 3DSCALE , XDS , X-GEN for X-ray single crystal diffraction data reduction. These non-graphical UIs are not intended to match the expert use of a particular program, but to provide a means to automatically process and characterize a data set, which includes determining Space Group, Resolution Cutoff, Rmerge, Completeness, Redundancy, I/SigI etc. They also provide a set of handy diagnostic tools to quickly identify problems, if any, which should be helpful for remote and/or quick data collection at synchrotron beamlines. Details of the various UI's and their application to real data will be presented. Work supported by the SER-CAT Member Institutions. References: 1). Otwinowski, Z. & Minor, W. (1997), Met. Enz. 276:307-326. 2). Pflugrath, J.W. (1999), Acta Cryst. D55:1718-1725. 3). Fu, Z.Q. (2005), Acta Cryst. D61:1643-1648. 4). Kabsch, W. (1988), J. Appl. Cryst. 21:916-924. 5). Howard, A.J. (1996), Proc.Macromol. Cryst. Comput. Sch., Oxford University Press,Oxford, UK.1

07.03.3 DrugSite: A Web-based Platform for Sharing Overlaid Protein:Ligand Complex Structures Barry Finzel, Ramprasad Akavaram, Aravind Ragipindi University of Minnesota College of Pharmacy, Minneapolis, MN, United States The popular drug design techniques of scaffold hopping and target hopping rely heavily on the accurate overlay of experimentally derived protein:ligand complex structures to emphasize both the differences and similarities in diverse examples of ligand binding. We have long believed that optimal superposition of such complexes results from an overlay of select protein substructures surrounding the binding site. The chosen substructure should include key structural elements contributing to complex stabilization, but not include elements subject to conformational change when different ligands bind. Suitable substructures are empirically identified, and the selection can evolve over the course of a project as different classes of ligands are added into the ensemble.

Modern software has not evolved to simplify this approach. Rather, the emphasis has been on ready access to algorithms employing sequence and secondary structure matching that seeks to empower users to overlay progressively more divergent protein family members. While these algorithms have their uses, they do not result in the most useful alignments of ligands for drug design applications.

We have been working to develop a database of overlay methods that captures optimized procedures for aligning important targets in drug-design. A web-based interface provides access to computational tools to apply these methods to either user-contributed complex structures or structures from the PDB, and allows users to share aligned structures with collaborators. The resulting platform provides an excellent means for communicating ligand structural data in a most useable form already aligned on relevant homolog structures for use by chemist and biochemist collaborators who might not otherwise have the expertise to devise an optimal superposition of the complexes.

07.03.4 Auto-Rickshaw: A tool for online validation of X-ray diffraction experiment and model completion Santosh Panjikar, Venkataraman Parthasarathy, Manfred Weiss, Victor Lamzin, Paul Tucker EMBL-Hamburg Outstation, Hamburg, Germany The Auto-Rickshaw is an automated structure determination software pipeline and it is based on several distinct computer coded crystallographic decision-makers, which invoke a variety of macromolecular crystallographic programmes/ programme packages during the structure determination process [1]. The primary aim of the pipeline is to validate the crystallographic experiment at the synchrotron site while the crystal is still at, or near, the beamline. The system is optimized for speed, so that structure determination can proceed within minutes after integrated and scaled diffraction data are available. Typically within a few minutes the answer is provided whether the collected data will be of sufficient quality to allow successful structure determination. Currently, the platform offers S/MAD, SIRAS, RIP and MR phasing protocols as well as combination of MR with various experimental phasing methods [2]. The result is an improvement of poor MR, or poor experimental phases and the determination of a larger percentage of the model, from X-ray data better than 2.9 resolution, in an automated manner. The platform has been installed on a Linux cluster at EMBL-Hamburg and is remotely accessible to the beamline users via a web-server [3]. It is accessible from most Internet browsers and allows beamline users to validate their X-ray diffraction experiments and model completion. Since 2008, Auto-Rickshaw web server [3] has been made accessible to the worldwide scientific community. References [1] Panjikar et al., (2005). Acta Cryst. D61, 449-457. [2] Panjikar et al., (2009). Acta Cryst. D65, 1089-1097 [3] http://www.embl-hamburg.de/Auto-Rickshaw/

07.03.5 The CCP4 Software Suite - Current Status and Future Developments Ronan Keegan , Martyn Winn , Eugene Krissinel , Charles Ballard , Natalie Zhao , George 1 Pelios STFC Rutherford Appleton Laboratory, Oxfordshire, United Kingdom, STFC Daresbury Laboratory, Cheshire, United Kingdom CCP4 exists to produce and support a world-leading, integrated suite of programs that allows researchers to determine macromolecular structures by X-ray crystallography. CCP4 aims to develop and support the development of cutting edge approaches to experimental determination and analysis of protein structure, and integrate these approaches into the suite. The current CCP4 software suite is on release series 6.1.x. A particular focus of these releases is the automation of significant parts of the structure solution process, including XIA2 for data processing, Crank for experimental phasing, MrBUMP and Balbes for Molecular Replacement, and Buccaneer for model building. There are also a number of new programs, including Pointless for Laue group and spacegroup determination, the new iMosflm interface, Parrot for density modification, and PISA for identification of protein-protein interfaces. We will give an overview of the additions to the CCP4 suite, as well as an update on established programs. A major overhaul of the CCP4 suite is under development. A new graphical front-end will provide easier control of the suite, and considerable help with interpreting and evaluating the results. At the core, there will be in-built support for automation, making straightforward structures simple to solve, while continuing support for more challenging projects. Finally, usage of the suite will be underpinned by better data management, with support for database back-ends. CCP4 also aims to enhance its functionality related to the maintenance and use of data on small molecules (ligands). Firstly, a considerably larger library of chemical compounds will be provided with the Suite. Extended search functions will be provided to allow for efficient retrieval of known compounds or their close analogs. Secondly, existing functions for generating restraint data for new ligands will be enhanced by the inclusion of relevant software, such as ProDRG, into the Suite, as well as by the development of new methods for structure reconstruction on the basis of partial similarity to structures in the library. Functionality will be available through a graphical front-end application, jLigand.1 2 1 2 1 1 1

07.03.6 How good is good enough? Predicting the success or failure of structure solution from first principles. James Holton University of California, San Francisco, CA, United States, Lawrence Berkeley Laboratory, Berkeley, CA, United States How much x-ray exposure is required to solve a structure? Multiplicity is good but how much will add too much read-out noise? What about a better detector? What about a perfect detector? Answering these questions requires that damage, noise and signal be placed on a common, absolute scale. To this end, a quantitative simulator of the entire diffraction experiment called "MLFSOM" (MOSFLM in reverse) was created. The input to the simulator is a protein data bank (PDB) file and parameters such as photon flux, crystal size and detector performance characteristics entered in conventional units such as photons/s and millimeters. MLFSOM was used to produce images in SMV format that were subsequently processed with ELVES. The general result of these trials was that one and only one of the many sources of noise in the diffraction experiment will dominate a given data set, and the optimal strategies for MAD/SAD and high-angle data collection are mutually exclusive. Faint, high-angle spots are best collected with exposures long enough to bury the detector read out noise under the background-photon noise (but no longer), but the optimal strategy for MAD/SAD was collecting a large number of very brief exposures, or dose slicing.1 2

07.04.1 Radiation damage in macromolecular crystallography: current challenges. Elspeth Garman , Ian Carmichael1 1 2 2

University of Oxford, Oxford, United Kingdom, Notre Dame Radiation Laboratory, Notre Dame, United States For protein crystals at room temperature, radiation damage during the diffraction experiment is rapid even on a laboratory X-ray source. In the past, the required data had to be collected from several different crystals and merged together. The intense X-ray beams produced by third generation synchrotrons can destroy crystalline order in a matter of seconds. Over the last 20 years, the use of cryo-cooling techniques which allow X-ray data to be collected with the sample held in a stream of cooled nitrogen gas at 100K, has become the norm [1, 2]; at 100K crystals can withstand many times the dose (J/kg=Gy) [3] compared with room temperature (depending on the dose rate [4]), and the necessary data can usually be obtained from a single crystal. However, observations of degradation of crystal diffraction with increasing radiation dose at 100K have now become commonplace at third generation synchrotrons. Researchers are trying to understand the physical and chemical processes involved in this damage (reviewed in [5,6]), which manifests itself in a number of different ways, including: changes in crystal colour, decreasing diffraction power with dose, a small but measurable linear increase in unit cell volume, and specific structural damage to covalent bonds in the amino acids of the protein molecules [7]. Enzyme active sites seem particularly sensitive to damage, so this phenomenon can lead to incorrect conclusions on biological mechanisms being drawn. Thus the issue of radiation damage during diffraction experiments has recently come to the fore as a concern for all structural biologists. Current issues being addressed and the challenges of research into this area will be outlined, informed by the material presented at the Sixth International Workshop on Radiation Damage to Biological Crystalline Samples held at SSRL in March 2010. References: [1] Teng, T-Y (1990) J Appl. Cryst. 23, 387-391 [2] Garman, E.F. & Schneider, TR (1997) J Appl. Cryst. (1997) 30, 211-237. [3] Owen, RL, Rudio-Piera, E and Garman, EF (2006) PNAS (2006) 103, 4912-4917. [4] Southworth-Davies, RJ.,Medina, MA.,Carmichael, I, & Garman, EF. Structure (2007) 15, 1341. [5] Ravelli, RGB & Garman, EF (2006) Current Opinion of Structural Biology (2006) 16, 624. [6] Garman, EF (2010) Acta Cryst. D66, 339-351. [7] Weik, M et al. (2000) PNAS 97, 623-628. [8] Burmeister, W.P. (2000) Acta Cryst. D56, 328-341. [9] Ravelli, R.G.B. & McSweeney, S. (2000) Structure 8, 315-328.

07.04.2 Quality Versus Quantity: the Role of Carefully Planned Diffraction Experiments in Highthroughput Crystallography Tobias Krojer, Frank von Delft Structural Genomics Consortium, Oxford, United Kingdom Our experience at SGC Oxford shows that the rate of success in a high-throughput environment does not only depend on the number of proteins going into crystallization, but on careful planning of data collection experiments. SGC-Oxford, is part of a world-wide structural genomics initiative and currently the Oxford site deposits four novel, human structures per month. Because our target list is fixed and the proteins consistently challenging, the emphasis in the crystallography group is on ensuring success even for marginal experiments through best practice data collection, which dramatically lowers the workload on the upstream pipeline. A vital ingredient in this philosophy is frequent access to high-quality beamlines, in our case at SLS and DIAMOND. Although these facilities provide equipment of unprecedented quality, we note that successful data collection still depends equally on experimenters' experience and skills. Here we present results from the numerous data sets that we have collected over the last six years at synchrotron beamlines and come up with suggestions for future software developments. We conclude that (i) beam sizes smaller than the diffracting volume of a crystal are of little benefit, and that (ii) exploratory datasets are of great value for predicting crystal lifetimes but hard to interpret for marginal diffraction. Thus, we still lack tools for routine experiments, specifically for characterizing the intersection of beam and crystal, as well as robust, real-time metrics for monitoring crystal decay.

07.04.3 The minimum crystal size needed to solve a structure. James Holton , Kenneth Frankel1 1,2 1 2

University of California, San Francisco, CA, United States, Lawrence Berkeley Laboratory, Berkeley, CA, United States The total amount of photons scattered into diffraction spots by a cryo-cooled protein crystal before it is dead is fixed because radiation damage and accumulated scattered intensity (photons/spot) are both proportional to fluence (incident photons/area). This means that damage-limited data quality is independent of data collection time, and therefore also independent of flux (photons/s). We calculated the damage-limited spot intensity from a protein crystal at a desired resolution given the molecular weight, crystal volume, solvent content, Wilson B factor and X-ray wavelength using classic scattering formulae and a simple spot-fading model. Theoretically, a perfect lysozyme crystal 1.2 micron in diameter should be sufficient for a complete data set (4 photons/hkl at 2 ), but background scattering on contemporary equipment pushes this minimum lysozyme size up to 8 microns. An easy-touse calculator for other cases is available at http://bl831.als.lbl.gov/xtalsize.html

07.04.4

Matthew Warkentin, Robert Thorne Cornell University, Ithaca, NY, United States We report the temperature dependence of global radiation damage to thaumatin crystals between T=300 and 100 K. The amount of damage for a given dose decreases sharply as the temperature decreases from 300 K to 220 K, and then decreases much more gradually on further cooling below the protein-solvent glass transition. We observe two regimes of temperature-activated behaviour. At temperatures above ~200 K, the activation energy of 4.3 kcal/mol indicates that radiation damage is dominated by diffusive motions: Diffusion of radicals through solvent channels, the diffusive relaxation of protein hydration water, and diffusive motions of loose side chains all have activation energies in this range. At temperatures below ~200 K the activation energy is only 0.24 kcal/mol, on the order of the thermal energy. Similar activation energies describe the temperature dependence of radiation damage to a large variety of small-molecule organic crystals over the temperature range between T=300 K and 80 K. These systems have atomic vibrational spectra and energies that are similar to those of proteins. This suggests that the temperature dependence of radiation damage below T=200 K is associated with the thermal occupation of the first few excited atomic vibrational states, and that diffusive processes do not contribute significantly to global damage. Below ~80 K, vibrational excitations are frozen out, zero point motions dominate, and global radiation damage becomes temperature independent. Using the radiation damage model of Blake and Phillips (1962), we show that radiation damage proceeds sequentially, with native protein first becoming disordered and then amorphous at all temperatures. The ratio of the amorphization rate to the disordering rate is constant below T~200 K but grows above it. Large scale conformational and molecular motions are frozen out below T=200 K, but become increasingly prevalent and make an increasing contribution to overall damage at higher temperatures. Blake, C., and Phillips, D.C. (1962). Effects of X-irradiation on single crystals of myoglobin. In Proceedings of the Symposium on the Biological Effects of Ionising Radiation at the Molecular Level (Vienna: International Atomic Energy Agency), pp. 183191.

07.04.5 Spatial dependence and mitigation of MX radiation damage by focusing Edward Stern , Yanhui Zou , Yizhak Yacoby , Andrzej Joachimiak , Randy Alkire , Kenneth 4 Evans-Lutterodt1 1 1 2 3 3

Univ. of Washington, Seattle, WA 98195, United States, Racah Institute of Physics, Hebrew 3 Univ., Jerusalem, Israel, Argonne National Laboratory, Argonne, IL 60439, United States, 4 Brookhaven National Lab, Brookhaven, NY, United States

2

Recently, strategies to reduce primary radiation damage have been proposed which depend on focusing x-rays to dimension smaller than the penetration depth of excited photoelectrons (PEs). For a line focus as used here the penetration depth is the maximum distance from the irradiated region along the x-ray polarization direction that the PEs penetrate. Reported here are measurements to determine the penetration depth and magnitude of PE damage excited by 18.6keV photons in a lysozyme crystal. It is found that the x-ray dose has a significant contribution from the crystals 9 w% solvent NaCl atoms. The 15.8 keV PEs of the Cl atoms and their accompanying 2.8 keV localized dose from the decay of the resulting excited atoms more than doubles the dose deposited in the focused region because of a much greater cross section and higher energy of the excited atom, degrading the mitigation of radiation damage. Eliminating heavier atoms from the solvent will significantly improve the mitigation of damage by focusing. The experimental results showed the penetration depth of ~17 keV PEs is 1.36+/- 0.2 m, well below previous theory estimates. Such a small penetration depth raises challenging technical issues to mitigate damage by focusing because the optimum requirements are gaussian line focused beams with sigma of 0.15 m and distance between lines of 1.8 m to reduced damage by a factor of 2.

07.04.6 Reduced Radiation Damage In Protein Crystals With Micron-Sized X-Ray Beams Robert F. Fischetti , Ruslan Sanishvili , Derek Yoder , Sudhir Pothineni , Janet L. Smith , 3 1 1 1 Gerold Rosenbaum , Shenglan Xu , Oleg Makarov , Sergey Stepanov , Venugopalan 1 1 Nagarajan , Stefan Vogt gm/Ca-Cat, Biosciences Division, Argonne National Laboratory, Argonne, Il, United States, life Scineces Inst., Dept. Of Biological Chemistry, U. Of Michigan, Ann Arbor, Mi, United 3 States, dept. Of Biochemistry And Molecular Biology, U. Of Georgia, Athens, Ga, United 4 States, experimentl Facilities Division, Aps, Argonne National Laboratory, Argonne, Il, United States2 1 1 1 1 1 1,2

Cryo-cooling of protein crystals significantly reduces X-ray induced radiation damage, but does not eliminate it. The predominant mechanism of interaction of an X-ray with an atom in the crystal is the emission of a photoelectron carrying most of the energy of the incident X-ray and causing damage as it deposits that energy in the crystal. When a photoelectron interacts with an atom, it loses energy slowly at first and then more rapidly as its energy decreases. Thus, if the beam size is small compared to the distance the photoelectron travels from its point of emission, then deposition of photoelectron energy outside the beam footprint may reduce radiation damage inside the beam footprint. Monte-Carlo simulations predict that a photoelectron of typical energy could travel 4 5 m from the point of emission before being absorbed. We studied radiation damage to lysozyme crystals by monitoring the diffracted intensity of 18.5-keV X-rays as a function of dose and beam size (0.86 20 m) at beamline 23-ID-B at the Advanced Photon Source. We observed a 3-fold reduction of damage per dose absorbed within the footprint of the smallest compared to the largest beam. In addition, the spatial extent of radiation damage was mapped using both 15.1- and 18.5-keV X-rays and a ~1- m beam. The damage profiles displayed spatial anisotropy with greater damage occurring along the direction of the X-ray polarization, as expected. The spatial extent of the damage was limited to about 4 m.

GM/CA CAT is supported by the NIH National Institute of General Medical Sciences and National Cancer Institute. The APS is supported by the US Department of Energy.

07.05.1 Probing reactions in real time using pair distribution function analysis Karena Chapman X-ray Science Division, Argonne, IL, United States The pair distribution function (PDF) method provides valuable insights into the local atomic structure in materials independent of crystallinity, heterogeneity or particle size. Recent advances in experimental methods and the advent of dedicated X-ray PDF beamlines, such as 11-ID-B at the Advanced Photon Source, have led to rapid growth in both PDF studies and the associated user community. This growth has occurred in parallel with the increasing interest in nanoscale and disordered materials, for which conventional Bragg crystallographic methods offer limited insight. Current state-of-the-art PDF set ups (with optimized beam intensity, sample environments and detectors) now allow total scattering data suitable for PDF analysis to be collected at up to 30 Hz. This allows for the structural changes during reactions to be probed in-situ to reveal changes in bonding during catalytic reactions and particle nucleation and growthfrom the earliest X-ray amorphous multi-atom clusters to nanoparticles and beyond. The insights gained into the reaction kinetics and mechanism can ultimately lead to greater control of structure and functional behavior.

07.05.2 STRUCTURE OF CRYSTALLOGRAPHICALLY STORAGE MATERIALS Hyunjeong Kim Los Alamos National Laboratory, Los Alamos, NM, United States Hydrogen is considered a promising alternative fuel for transportation, provided we can find a way to store a large amount of it in a compact way. The realization of such a storage system can be achieved by developing materials that can easily absorb, safely store, and rapidly release hydrogen. However, there is currently no material to meet all the requirements for on board storage. Great efforts have been made to look for a way to improve properties or to prepare new materials. One of widely adopted ways to prepare new hydrogen storage materials is mechanical alloying or ball milling. Materials prepared by this method are often nano- or amorphous-phases or mixture of both and they exhibit interesting hydrogen storing properties [1]. Alternatively, packing hydrogen storage materials into porous materials [2] leads to great improvement in their properties; such nano-confinement allows materials to release high purity hydrogen at lower temperatures without a significantly long induction period. Despite of favorable changes in properties, little is known about the structure of both types of systems. This is partly because amorphous or nano-sized nature limits the use of conventional crystallographic analysis and, therefore, structural determination becomes very challenging. In this talk, I will present our local structural studies on such crystallographically challenged hydrogen storage materials by using the atomic pair distribution function analysis [3] on total scattering data. The systems of interest are MgxCo100-x alloys prepared by ball milling [4] and nano-phase ammonia borane (NH3BH3) confined in pores of mesoporous silica MCM-41 [5]. [1] S. Orimo and H. Fujii, Appl. Phys. A 72, 167-186 (2001). [2] A. Gutowska et al., Angew. Chem. Int. Ed. 44, 3578-3582 (2005); R. K. Bhakta et al., J. Am. Chem Soc. 131, 13198-13199 (2009). [3] T. Egami & S. J. L. Billinge, Underneath the Bragg Peaks: Structural Analysis of Complex Materials, Pergamon Press Elsevier, Oxford, England, 2003; Th. Proffen & H. J. Kim, J. Mater. Chem. 19, 5078-5088 (2009). [4] Y. Zhang et al., J. Alloys Compd. 393, 147-153 (2005); H. Shao et al., Scripta Materialia 60, 818-821 (2009); J. Matsuda et al., Nanotechnology 20, 204015 (2009). [5] H. J. Kim et al., J. Am. Chem. Soc. 131, 13749-13755 (2009). CHALLENGED HYDROGEN

07.05.3 Local Structure Effects in Magnetoresistance Materials Efrain Rodriguez , Anna Llobet , Thomas Proffen , Katharine Page , Mark Green1 2 1 2 2 2 1

NIST Center for Neutron Research, Gaithersburg, MD, United States, Lujan Neutron Center, LANL, Los Alamos, NM, United States The magnetoresistance (MR) effect is a technologically important property employed in magnetic hard disks and sensors. In efforts to find new materials with promising MR properties, researchers have focused on the mixed valence manganites Ln1-xAxMnO3 where Ln is a trivalent lanthanide and A is a divalent alkaline earth metal. Another set of materials showing promise in this area is the series of solid solutions with the stoichiometries Zn1xCuxCr2Se4. This talk will focus on how we employed neutron and X-ray powder diffraction to obtain the local and long-range structure of both materials to have a better understanding of the microscopic interactions leading to the MR effect. In particular, we present the case of La0.5Ca0.5MnO3 below the charge-ordering temperature. By combining Rietveld and pair distribution function (PDF) analysis with the total neutron scattering data, we examined two competing models describing the low temperature, charge-ordered/orbital-ordered (CO-OO) 3+/ 4+ phase: 1) the Mn Mn checkerboard model and 2) the Mn-Mn dimer model or so-called Zener polaron model. In the case of the selenides, we use PDF analysis of X-ray data to find how the local environment of the Cu and Cr cations lead to the observed magnetic and transport properties.

07.05.4 Coupling total scattering and density functional theory computations to solve the structure of complex disordered aluminosilicates Claire White , John Provis , Thomas Proffen , Daniel Riley , Jannie van Deventer1 2 1 1 2 1 1

University of Melbourne, Victoria, Australia, Los Alamos National Laboratory, Los Alamos, NM, United States Understanding the atomic structure of complex metastable materials is of great importance in research and industry, however, such structures resist solution by most standard techniques. Here, a novel synergy between total scattering and density functional modelling is presented to solve the structure of the metastable aluminosilicate material metakaolin. Metakaolin is obtained by calcination of kaolinite, and is used in quantities of millions of tonnes per annum in blending with Portland cement for concretes, as well as being a useful geopolymer precursor, and is a key intermediate in processing of many fired ceramics. The structure is elucidated by two independent methods, both based on the combination of total scattering/pair distribution function analysis (PDF) and density functional theory (DFT).

In the first method, the structure is obtained by iteration between least-squares real-space refinement using neutron PDF data, and geometry optimisation using DFT. The resulting structural representation is both energetically feasible and in excellent agreement with experimental data. In the second, the process of kaolinite dehydroxylation is modeled using DFT and a step-wise methodology, where several water molecules at a time are removed from the original kaolinite structure, geometry optimization is carried out, and the process is repeated until the dehydroxylated structure is reached. The structures generated during the dehydroxylation process are then validated by comparison with X-ray and neutron PDF data.

This study provides new insight into the local environment of the aluminum atoms in metakaolin, including evidence of the existence of tri-coordinated aluminum. By the availability of this detailed atomic description of its structure, there exists the opportunity to tailor chemical and mechanical processes involving metakaolin and other complex metastable materials at the atomic level to obtain optimal performance at the macro-scale.

07.05.5 Phase progression of alumina nanoparticle catalyst supports as a function of synthetic temperature Stacey Smith , Branton Campbell , Baiyu Huang , Calvin Bartholomew , Brian Woodfield , 2 4 4 5 Juliana Boerio-Goates , Katherine Page , Hyunjeong Kim , Karena Chapman Brigham Young University, Physics & Astronomy, Provo, UT, United States, Brigham Young 3 University, Chemistry & Biochemistry, Provo, UT, United States, Brigham Young University, 4 Chemical Engineering, Provo, UT, United States, Los Alamos National Laboratory, Los 5 Alamos Neutron Science Center, Los Alamos, NM, United States, Argonne National Laboratory, Advanced Photon Source, Chicago, IL, United States We have developed a simple and uniquely cost-effective synthetic method for producing Al2O3 nanoparticles of exceptional size (3-5 nm) and purity. The product shows promise as an improved industrial catalyst support due its enhanced surface area and the mesoporous character of its agglomerates. To establish the temperature range through which we can produce the catalytically-active gamma phase, we must determine the phase progression of our samples as a function of synthetic temperature. This is challenging because the alumina phase diagram includes many closely-related phases that are not readily distinguished from powder-diffraction data due to the extremely particle-size broadened Q-space peaks. In these cases, PDF analysis was able to resolve the distinct local structures of the candidate phases. We will demonstrate that a combination of PDF and Rietveld refinements best resolves our alumina nanoparticle phase progression pathway.1 2 1,2 1 2 3 2

07.05.6 Ferroelectric-relaxor crossover in Ba(Ti1 xZrx)O3 studied using neutron total scattering measurements and reverse Monte Carlo modeling Ilkyoung Jeong, C. Y. Park, J. S. Ahn, S. Park, D. J. Kim Pusan National University, Busan, Korea, Republic of Comprehensive structural studies on normal ferroelectric to relaxor crossover in Ba(Ti1-xZrx)O3 (BTZ) are performed using neutron total scattering measurements analyzed by reverse Monte Carlo modeling. In BTZ solid solution, we estimated the degree of the displacement correlation between Ti ions and found that it is stronger and extends much longer for ferroelectric state than relaxor state. In addition, we present evidence that the overall o_centering behavior of Ti ion changes from directional to random displacements between ferroelectric and relaxor phases, and thus provide atomistic picture for ferroelectric-relaxor crossover with increasing Zr concentration.

07.05.7 Applications of Single Crystal Diffuse X-ray Scattering for Studies of Polymorphism in Pharmaceuticals. Eric Chan, Darren Goossens, Aidan Heerdegen, Richard Welberry Australian National University, Canberra, Australia Due to the recent advancements in modern computing power, the analysis and interpretation of single crystal X-ray diffuse scattering for molecular crystals now involves the construction of a computer model of a dynamic crystal. The method allows inclusion of structural features on a local level that may be tested against their effect on the observed diffuse scattering. This gives great insight into the dynamic behavior of organic molecules in the solid state and the models can also be used to explain the structural nature of packing defects or lattice strain. In this paper we discuss the analysis of three polymorphic systems; namely, benzocaine, paracetamol and aspirin. For benzocaine, a low temperature phase transition occurs whereby the orthorhombic phase (form II) transforms to a twinned monoclinic phase (form III). The low temperature twinned crystal displays many 'well-defined' Bragg peaks. For the room temperature form II, diffuse scattering features are observed in the absence of 'low temperature' Bragg peaks which, when modeled, show that at a local level the form II crystal has a structure which exhibits precursor effects of the incipient phase transition. All the simulations use Hooke's law springs associated with intermolecular connections to approximate the normal modes of vibration in a molecular crystal. For benzocaine a simplified set of important connections were used and force constants needed to be determined through trial and error. The work on the monoclinic and orthorhombic forms of paracetamol demonstrates that much trial and error is no longer necessary and the force constants in a model can be approximated from knowledge of Van der Waals radii provided that all intermolecular interactions within a certain distance threshold are taken into consideration. The approximation works well and its effect on the simulation is shown quantitatively using a least squares refinement. Results from modeling diffuse data collected from form II of aspirin suggests that the crystal has undergone layer dislocations, during or after the crystallization, that resemble the form I packing. Because these layer dislocations are not perfect within the crystal, a resultant lattice strain is also observed in the diffuse scattering. This strain should affect the solid-state physical properties of the form II crystal.

07.06.1 Precise Absolute Structure Refinement Using Quotient Restraints Simon Parsons , Howard Flack , Oliver Presley , Trixie Wagner , Paul McGovern1 2 1 2 3 4 1

University of Edinburgh, Edinburgh, Scotland, United Kingdom, University of Geneva, 3 4 Geneva, Switzerland, Oxford Diffraction, Yarnton, Oxfordshire, United Kingdom, Novartis, Basel, Switzerland In an absolute structure determination one absolute structure is refined competitively against the inverted alternative. The result is expressed by the Flack parameter x(u), which for absolute configuration determination can be interpreted as the mole fraction of the alternative enantiomer in the crystal. The physical range of x is 0 to 1 and even if the bulk material is known to be enantiopure the standard uncertainty (u) should be less than 0.1 before any firm conclusions regarding the absolute structure can be drawn. This criterion has proved to be extremely demanding for light-atom structures.

We have developed a method for absolute structure determination based on the quantity

D(h)

I (h) - I (-h) I (h) I (-h)

(1- 2 x )

F 2 (h) - F 2 (-h) F 2 (h) F 2 (-h)

The term based on I(h) and I(-h) and its standard uncertainty can be calculated from a single 2 2 crystal X-ray diffraction data set. The term based on F (h) and F (-h) can be calculated from the model. It is therefore possible to write out a set of restraints based on observed and calculated values of D(h) and apply these in an absolute structure refinement. Systematic errors in the intensities, such as absorption, tend to cancel out (in an average way) so that measured values of D(h) should be more accurate that the values of the measured intensities.

We have found that this method yields significantly more precise values of the Flack parameter than conventional refinement. For example when a data set was collected for Lalanine with Cu-K radiation at 100 K, conventional refinement yielded a Flack parameter equal to 0.12(21), whereas the restrained refinement yielded a value of 0.00(8). The method also carries the advantage that the Flack parameter is allowed to refine along with all the other parameters, so that its standard uncertainty reflects correlations present in the refinement.

07.06.2 Experimental conditions for absolute structure determination using Bayesian statistics Martin Lutz , Rob W. W. Hooft1 1 2 2

Utrecht University, Utrecht, Netherlands, Netherlands Bioinformatics Centre, Nijmegen, Netherlands The first absolute structure determination of an organic molecule was performed by Bijvoet and coworkers based on intensity differences of 15 pairs of reflections. This approach was followed by many similar studies, often with a slightly different way to select the reflection pairs or with different weighting of the selected pairs. In the 1980s the original procedure of examining a subset of Bijvoet pairs was superseded by the inclusion of an absolute structure parameter in the least-squares refinement. A renaissance of the Bijvoet method appeared by a contribution of Hooft, Straver and Spek (2008), where likelihood calculations in combination with Bayesian statistics are applied. In contrast to the original Bijvoet method, where only a subset of reflection pairs is considered, the Hooft method takes all Bijvoet pairs into account. The Hooft method appears to be very successful, even if only weak anomalous scatterers (e.g. oxygen) are present. This paper will deal with the experimental conditions, which are necessary for a reliable absolute structure determination. Special emphasis will be on the standard uncertainties of the experimental intensities. Outlier handling will be discussed on the assumption of a normal error distribution. In case the error distribution is non-Gaussian, use is made of the Student tdistribution to increase the robustness of the method. Example data are taken from our own laboratory as well as from Acta Crystallographica, where reflection data are deposited as supplementary material.

07.06.3 Improved Light-Atom Absolute Configuration Determination Using Sources.1 2 1 1

Microfocus1

Michael Ruf , Holger Ott , Matthew M. Benning , Bruce C. Noll , Charles F. Campana1

Bruker AXS Inc., Madison, Wisconsin, United States, Germany

2

Bruker AXS GmbH, Karlsuhe,

Determination of absolute configuration for light-atom structures has become central to research in pharmaceuticals and natural products synthesis. In the absence of elements heavier than silicon, it is often problematic to make a significant assignment of absolute configuration. Traditionally, a heavy-atom derivative has been prepared, but this is not always feasible. Making these assignments has become somewhat easier with the advent of highintensity microfocus sources, as the increased flux density can improve the anomalous signal from these samples through improvements in counting statistics. The improvement in data quality from a high-intensity microfocus source will be demonstrated in comparison to data from a conventional sealed-tube source.

07.06.4 Absolute Structure Determination - Interpreting the Flack Parameter Amber L. Thompson, David J. Watkin Chemical Crystallography, Oxford, United Kingdom It is now well established that the different enantiomers of a chiral material can have significantly different physiological properties - for example d and l-limonene. As a consequence of this, drug manufacturers and drug authorisation authorities are increasingly concerned about the absolute configuration of active pharmaceutical ingredients. In appropriate cases, X-ray structure analysis can give very reliable results.

The first absolute structure determination, of sodium rubidium tartrate, was carried out in 1951 by Bijvoet, Peerdeman and van Bommel. Until Rogers introduced his eta parameter into the least-squares refinement (1981), direct comparisons of Bijvoet pairs, or the application of the Hamilton R-factor ratio test were the principal crystallographic techniques used to assign absolute structures. Rogers eta parameter was quickly superseded by the Flack "x" parameter, a least-squares parameter which treated the crystal as a mixture of the original enantiomer and its twin by inversion. Flack pointed out that where as the Rogers parameter (which varied between +1 and -1) had no physical meaning as it approached the mid point, zero, the Flack parameter had a physical meaning over its entire range (from 0 to 1). A Flack parameter somewhere near the middle of the range, and with a suitable small e.s.d., indicated that the sample was twinned by inversion.

The incorporation of this parameter into most refinement programs, its ease of use, and its apparent robustness to less-than-ideal data collection strategies contributed to its rapid acceptance, and to misunderstandings about its interpretation. In 2000, almost 20 years after the Flack parameter was first described, Flack and Bernardinelli described the statistical interpretation of the parameter. In spite of this, there continued to be a hunch amongst practical crystallographers that Flack's own interpretation of his parameter was unduly pessimistic. The publication of the derivation of the Hooft, Straver and Spek parameter in 2008 encouraged us to carry out a critical analysis of 15