Edwin M. Westbrook

The Structural Biology Center 19ID undulator beamline: facility specifications and protein crystallographic results

The 19ID undulator beamline of the Structure Biology Center has been designed and built to take full advantage of the high flux, brilliance and quality of X-ray beams delivered by the Advanced Photon Source. The beamline optics are capable of delivering monochromatic X-rays with photon energies from 3.5 to 20 keV (3.5-0.6 A wavelength) with fluxes up to 8-18 x 10(12) photons s(-1) (depending on photon energy) onto cryogenically cooled crystal samples. The size of the beam (full width at half-maximum) at the sample position can be varied from 2.2 mm x 1.0 mm (horizontal x vertical, unfocused) to 0.083 mm x 0.020 mm in its fully focused configuration. Specimen-to-detector distances of between 100 mm and 1500 mm can be used. The high flexibility, inherent in the design of the optics, coupled with a kappa-geometry goniometer and beamline control software allows optimal strategies to be adopted in protein crystallographic experiments, thus maximizing the chances of their success. A large-area mosaic 3 x 3 CCD detector allows high-quality diffraction data to be measured rapidly to the crystal diffraction limits. The beamline layout and the X-ray optical and endstation components are described in detail, and the results of representative crystallographic experiments are presented.

[17] Charge-coupled device-based area detectors.

Publisher Summary This chapter discusses the physical and geometric properties of charge-coupled device (CCD) detector designs and examines the way in which a system without image amplification, despite its low signal level, can operate efficiently and with excellent sensitivity. Photon-counting detectors include traditional scintillation counter-based diffractometers and multiwire proportional gas chambers. Counting detectors must discriminate each X-ray photon—an electronic process requiring a finite time (its dead time). Integrating detectors have virtually no upper rate limits because they measure the total energy deposited during an integration period (although individual pixels can become saturated if the signal exceeds its storage capacity). The physical and electronic properties of a charge-coupled device are discussed in the chapter. The Advanced Photon Source (APS)-1 detector is a significant part of an integrated technical facility that also features an undulator X-ray source on the APS, carefully designed X-ray optics, modern networking, and a powerful computing environment (multiple processors, fast, large RAID disk arrays).

MAD analysis of FHIT, a putative human tumor suppressor from the HIT protein family

BACKGROUND The fragile histidine triad (FHIT) protein is a member of the large and ubiquitous histidine triad (HIT) family of proteins. It is expressed from a gene located at a fragile site on human chromosome 3, which is commonly disrupted in association with certain cancers. On the basis of the genetic evidence, it has been postulated that the FHIT protein may function as a tumor suppressor, implying a role for the FHIT protein in carcinogenesis. The FHIT protein has dinucleoside polyphosphate hydrolase activity in vitro, thus suggesting that its role in vivo may involve the hydrolysis of a phosphoanhydride bond. The structural analysis of FHIT will identify critical residues involved in substrate binding and catalysis, and will provide insights into the in vivo function of HIT proteins. RESULTS The three-dimensional crystal structures of free and nucleoside complexed FHIT have been determined from multiwavelength anomalous diffraction (MAD) data, and they represent some of the first successful structures to be measured with undulator radiation at the Advanced Photon Source. The structures of FHIT reveal that this protein exists as an intimate homodimer, which is based on a core structure observed previously in another human HIT homolog, protein kinase C interacting protein (PKCI), but has distinctive elaborations at both the N and C termini. Conserved residues within the HIT family, which are involved in the interactions of the proteins with nucleoside and phosphate groups, appear to be relevant for the catalytic activity of this protein. CONCLUSIONS The structure of FHIT, a divergent HIT protein family member, in complex with a nucleotide analog suggests a metal-independent catalytic mechanism for the HIT family of proteins. A structural comparison of FHIT with PKCI and galactose-1-phosphate uridylyltransferase (GaIT) reveals additional implications for the structural and functional evolution of the ubiquitous HIT family of proteins.

CCD-based detector for protein crystallography with synchrotron X-rays☆

A detector with a 114 mm aperture, based on a charge-coupled device (CCD), has been designed for X-ray diffraction studies in protein crystallography. The detector was tested at the National Synchrotron Light Source with a beam intensity, through a 0.3 mm collimator, of greater than 10(9) X-ray photons/s. A fiberoptic taper, an image intensifier, and a lens demagnify, intensify, and focus the image onto a CCD having 512 x 512 pixels. The statistical uncertainty in the detector output was evaluated as a function of conversion gain. From this, a detective quantum efficiency (DQE) of 0.36 was derived. The dynamic range of a 4 x 4 pixel resolution element, comparable in size to a diffraction peak, was 10(4). The point-spread function shows FWHM resolution of approximately 1 pixel, where a pixel is 160-mu-m on the detector face. A data set collected from a chicken egg-white lysozyme crystal, consisting of 495 0.1-degrees frames, was processed by the MADNES data reduction program. The symmetry R-factors for the data were 3.2-3.5%. In a separate experiment a complete lysozyme data set consisting of 45 1-degrees frames was obtained in just 36 s of X-ray exposure. Diffraction images from crystals of the myosin S1 head (a = 275 angstrom) were also recorded; the Bragg spots, only 5 pixels apart, were separated but not fully resolved. Changes in the detector design that will improve the DQE and spatial resolution are outlined. The overall performance showed that this type of detector is well suited for X-ray scattering investigations with synchrotron sources.

CCD sensors in synchrotron X-ray detectors

Abstract The intense photon flux from advanced synchrotron light sources, such as the 7-GeV synchrotron being designed at Argonne, require integrating-type detectors. Charge-coupled devices (CCDs) are well suited as synchrotron X-ray detectors. When irradiated indirectly via a phosphor followed by reducing optics, diffraction patterns of 100 cm 2 can be imaged on a 2 cm 2 CCD. With a conversion efficiency of ∼ 1 CCD electron/X-ray photon, a peak saturation capacity of > 10 6 X-rays can be obtained. A programmable CCD controller operating at a clock frequency of 20 MHz has been developed. The readout rate is 5 × 10 6 pixels/s and the shift rate in the parallel registers is 10 6 lines/s. The test detector was evaluated in two experiments. In protein crystallography diffraction patterns have been obtained from a lysozyme crystal using a conventional rotating anode X-ray generator. Based on these results we expect to obtain at a synchrotron diffraction images at a rate of ∼ 1 frame/s or a complete 3-dimensional data set from a single crystal in ∼ 2 min. In electron energy-loss spectroscopy (EELS), the CCD was used in a parallel detection mode which is similar to the mode array detectors are used in dispersive EXAFS. With a beam current corresponding to 3 × 10 9 electron/s on the detector, a series of 64 spectra were recorded on the CCD in a continuous sequence without interruption due to readout. The frame-to-frame pixel signal fluctuations had σ = 0.4% from which DQE = 0.4 was obtained, where the detector conversion efficiency was 2.6 CCD electrons/X-ray photon. These multiple frame series also showed the time-resolved modulation of the electron microscope optics by stray magnetic fields.

Characterization and data collection on a direct-coupled CCD X-ray detector

Abstract A large-area, multi-module, CCD-based detector without intensification stages is being developed by our group for X-ray diffraction applications. Each module consists of a fiberoptic taper with a phosphor deposited on the large end and a large-format, scientific CCD bonded to the small end. A single module has been constructed to evaluate the performance of this type of detector. This module has an active area of 43 × 43 mm2, a point response function FWHM = 80 μm, a dynamic range of > 20 000:1, and a high DQE. Using four parallel readout circuits, the CCD can be read out in 1.8 s. Crystallographic data collected using a rotating-anode source demonstrate the capability of this type of detector.

CCD-based synchrotron x-ray detector for protein crystallograph-performance projected from an experiment

The intense x radiation from a synchrotron source could, with a suitable detector, provide a complete set of diffraction images from a protein crystal before the crystal is damaged by radiation (2-3 min). An area detector consisting of a 40 mm dia. x-ray fluorescing phosphor, coupled with an image intensifier and lens to a CCD image sensor, was developed to determine the effectiveness of such a detector in protein crystallography. The detector was used in an experiment with a rotating anode x-ray generator. Diffraction patterns from a lysozyme crystal obtained with this detector are compared to those obtained with film. The two images appear to be virtually identical. The flux of 10/4 x-ray photons/s was observed on the detector at the rotating anode generator. At the 6-GeV synchrotron being designed at Argonne, the flux on an 80×80 mm2 detector is expected to be ≫ 109 photons/s. The projected design of such a synchrotron detector shows that a diffraction-peak count ≫ 106 could be obtained in ˜0.5 s. With an additional ˜0.5 s readout time of a 512×512 pixel CCD, the data acquisition time per frame would be ˜1 s so that ninety 1o diffraction images could be obtained, with approximately 1% precision, in less than 3 min.

Area detector design Part II. Application to a modular CCD-based detector for X-ray crystallography

Abstract The performance of laboratory and synchrotron CCD-based detectors for X-ray crystallography is modeled using expressions which describe both the detector and the experiment. The detectors are constructed from an array of identical modules, each module consisting of a phosphor X-ray-to-light convertor, a fiberoptic taper and a CCD. The performance is characterized by the detective quantum efficiency (DQE) and dynamic range (DR), and by four additional expressions; the detective collective efficiency (DCE), experimental detective quantum efficiency (XDQE), experimental detective collection efficiency (XDCE) and experimental dynamic range (XDR). These additional expressions provide a means for including experimental constraints in the design of the detector. For a crystallography detector, these constraints include the requirements that the detector a) integrate Bragg peaks to maximum precision, and b) efficiently collect data to high resolution (large Bragg angle). Results obtained using these expressions demonstrate the need for a detector with a relatively large area. In order to build a such a detector from a reasonable number of modules using currently available fiberoptic tapers and CCDs, tapers with a demagnification ratio of > 3:1 are required. A different conclusion would be arrived at if the DQE alone were considered, demonstrating the importance of this method.

Crystal structure of the global regulator FlhD from Escherichia coli at 1.8 A resolution.

FlhD is a 13.3 kDa transcriptional activator protein of flagellar genes and a global regulator. FlhD activates the transcription of class II operons in the flagellar regulon when complexed with a second protein FlhC (21.5 kDa). FlhD also regulates other expression systems in Escherichia coli. We are seeking to understand this plasticity of FlhDs DNA‐binding specificity and, to this end, we have determined the crystal structure of the isolated FlhD protein. The structure was solved by substituting seleno‐methionine for natural sulphur‐methionine in FlhD, crystallizing the protein and determining the structure factor phases by the method of multiple‐energy anomalous dispersion (MAD). The FlhD protein is dimeric. The dimer is tightly coupled, with an intimate contact surface, implying that the dimer does not easily dissociate. The FlhD monomer is predominantly α‐helical. The C‐termini of both FlhD monomers (residues 83–116) are completely disrupted by crystal packing, implying that this region of FlhD is highly flexible. However, part of the C‐terminus structure in chain A (residues 83–98) was modelled using a native FlhD crystal. What is seen in chain A suggests a classic DNA‐binding, helix–turn–helix (HTH) motif. FlhD does not bind DNA by itself, so it may be that the DNA‐binding HTH motif becomes rigidly defined only when FlhD forms a complex with some other protein, such as FlhC. If this were true, it might explain how FlhD exhibits plasticity in its DNA‐binding specificity, as each partner protein with which it forms a complex could allosterically affect the binding specificity of its HTH motif. A disulphide bridge is seen between the unique cysteine residues (Cys‐65) of FlhD native homodimers. Alanine substitution at Cys‐65 does not affect FlhD transcription activator activity, suggesting that the disulphide bond is not necessary for either dimer stability or this function of FlhD. Electrostatic potential analysis indicates that dimeric FlhD has a negatively charged surface.

Using computational grid capabilities to enhance the capability of an X-ray source for structural biology

The Advanced Photon Source at Argonne National Laboratory enables structural biologists to perform state-of-the-art crystallography diffraction experiments with high-intensity X-rays. The data gathered during such experiments is used to determine the molecular structure of macromolecules to enhance, for example, the capabilities of modern drug design for basic and applied research. The steps involved in obtaining a complete structure are computationally intensive and require the proper adjustment of a considerable number of parameters that are not known a priori. Thus, it is advantageous to develop a computational infrastructure for solving the numerically complex problems quickly, in order to enable quasi-real-time information discovery and computational steering. Specifically, we propose that the time-consuming calculations be performed in a “computational grid” accessing a large number of state-of-the-art computational facilities. Furthermore, we envision that experiments could be conducted by researchers at their home institution via remote steering while a beamline technician performs the actual experiment; such an approach would be cost-efficient for the user. We conducted a case study involving multiple tasks of a structural biologist, including data acquisition, data reduction, solution of the phase problem, and calculation of the final result - an electron density map, which is subsequently used for modeling of the molecular structure. We developed a parallel program for the data reduction phase that reduces the turnaround time significantly. We also distributed the solution of the phase problem in order to obtain the resulting electron density map more quickly. We used the GUSTO testbed provided by the Globus metacomputing project as the source of the necessary state-of-the-art computational resources, including workstation clusters.