Imaging of cellular architecture and, ultimately imaging of macromolecular complexes and individual proteins, particularly membrane proteins, within a cellular environment, is a "holy grail" of cell and molecular biology. For example, 20 to 25% of all proteins are membrane-embedded and this class of proteins represents approximately 70% of all known pharmaceutical drug targets. Elucidation of the structures of membrane proteins would greatly facilitate the rational design of novel drugs to target conditions such as inflammation, depression, and pain. However while soluble proteins are readily crystallized and studied by X-ray crystallography (about 25,000 structures are available at 2-3 Å resolution), membrane proteins are much more difficult to crystallize and only a dozen have been determined to atomic resolution. A few hundred structures are available, mostly by electron microscopy (EM) techniques at a resolution of about 7 Å or worse. This Centre will attempt to determine structures of membrane proteins (and sub-micron protein crystals which are routinely produced in crystal screens) that are refractory to current X-ray techniques, and enable X-ray diffraction data to be collected from samples currently only being obtained from cryo-EM techniques.

Thus the high-resolution imaging techniques that this Centre aims to develop have the potential to revolutionise molecular biosciences. They could turn our current 'cartoon-sketch' level of understanding of membrane proteins into detailed pictures. While the ultimate application of these techniques will span many areas of research, we will here target a limited number of outstanding issues as test-bed systems for cellular and sub-cellular imaging and structural determination.

Cellular Imaging: Our test system for cellular imaging is the malaria parasite-infected erythrocyte. The malaria parasite resides for part of its lifecycle inside human red blood cells. The parasite establishes a system of membranes within the host cell cytosol and induces changes in the host cell membrane. These changes are critical to the disease pathology and lead to more than two million deaths per year. The cellular architecture of infected erythrocytes has previously been studied by electron and optical microscopy17, . EM allows excellent resolution of the cellular architecture but requires extensive fixing and processing that may cause alterations in the organisation of membranous structures. Fluorescence microscopy allows visualisation of fluorescently-labelled components in infected erythrocytes but has limited resolution. Because of the limitations of the available techniques, there is on-going debate regarding the origin and organisation of the different membranous structures in the host cell cytoplasm17, . Soft X-rays have wavelengths of about 1-10 nm. These wavelengths allow imaging at high spatial resolution and intracellular structures of P. falciparum infected erythrocytes have been successfully imaged by soft X-ray microscopy , at the 40-50 nm level. In this work, we propose to use coherent diffractive imaging methods to obtain additional information about cellular architecture of parasitised erythrocytes at a resolution of 10 nm or better. We propose to generate transfected malaria parasites expressing chimeras of proteins targeted to different compartments in the infected erythrocytes using established techniques , . Cells expressing recombinant proteins containing a six histidine tag will be incubated with KMnO₄. The resultant Mn⁴⁺ ions will bind to the polyhistidine sequences, forming a dense precipitate that strongly diffracts X-rays11. Images collected using the X-ray techniques will be compared with images obtained by electron microscopy and optical imaging17,18.

Subcellular imaging: We propose to examine important aspects of mitochondrial organisation as a test of our ability to image sub-cellular structures in mammalian cells. In most cells, mitochondria form a dynamic, branched reticulum that extends throughout the cytoplasm - . Their diversity in shape reflects a multiplicity of roles in cell development and differentiation . Alterations in mitochondrial morphology often result as a secondary consequence of disruptions in mitochondrial function, which includes a wide variety of important cellular processes, including energy generation, metabolism, cell death, and aging - . The high resolution diffractive imaging methods to be developed in the Centre have the potential to unravel some of the questions underlying m mitochondria fusion and fission. Determining the structures of macromolecular complexes: As an initial test of our capabilities for imaging macromolecular complexes we will examine the structures of amyloid fibrils that are associated with diseases such as Alzheimer's and Bovine Spongiform Encephalopathy. Amyloid fibrils have dimensions of 10-20 nm by up to several m . Interestingly some malaria proteins can also form amyloid-like fibrils . We will also examine the structure of Complex I (NADH-Ubiquinone oxidoreductase) of the inner mitochondrial membrane, one of the largest known membrane protein complexes. In humans, Complex I contains 46 different subunits. Defects in Complex I are the major cause of inherited disorders of mitochondrial energy generation, the most severe causing infant death. Other defects present later in life causing a range of degenerative diseases, particularly affecting brain, muscle and heart. Defects in Complex I have also been observed in Parkinson's disease, Alzheimer's disease and diabetes24. Despite its importance, the way in which Complex I functions in electron transfer and proton movement is poorly understood. Unlike other respiratory complexes, the structure of Complex I at atomic resolution is not known as generation of crystals suitable for X-ray diffraction has not been achieved. Owing to its large size and detergent extractability, Complex I can be purified quite easily from bovine heart mitochondria in large amounts. Complex I therefore serves as an excellent target for structural elucidation using the methods being developed here. We propose to generate 3-D microcrystals and 2-D arrays of this protein complex and to use our high energy X-ray imaging techniques to provide structural information about this complex. As an additional test system for structure determination, we propose to image a translocase complex that is critical for mitochondrial function30. A complex between the channel protein, Tom40, the receptor, Tom22, and a small subunit, Tom7, can be extracted with detergent from human cells. While much research has focussed on characterising this essential translocation apparatus, important knowledge is missing due to the lack of structural detail at a molecular level. A long term goal of CI Ryan and PI Gulbis is to solve the structure of the protein subunits separately and in complex with one another. Current work involves the production of recombinant proteins in E. coli as domains or whole proteins and purification complex directly from bovine liver mitochondria for crystallisation trials. For the X-ray test system, this complex will be prepared as microcrystals or reconstituted as an ordered 2-D semi-crystalline lattice within the plane of a synthetic membrane environment. The 2-D crystals will be soaked with heavy atom derivatives and used with X-ray free electron lasers to generate structural information on these proteins. These studies will provide important mechanistic insights into the molecular basis for the translocation of proteins across membranes.

Sample preparation: Methods of sample preparation developed for cryo-electron microscopy will be used for accurate 3-D reconstruction of macromolecules. With this technique, the sample is typically applied to a grid coated with holey carbon film and rapidly plunged into liquid ethane. This results in the formation of vitrified ice, and the sample is preserved with a minimum of structural degradation. The best resolution obtained for a single molecule structure by cryo EM and single particle reconstruction is currently 7.4 Å. For 2-D structures (for proof of principal we shall use currently available 2-D crystals of bacterial rhodopsin) the resolution that is achievable using X-ray diffraction will be affected by absorption of the vitrified ice support and distortions in the lattice due to lack of rigidity or the array laying on a distorted surface give rise to loss of higher resolution Bragg electron diffraction. This may be compensated to some extent by computer processing but the thickness of film containing the sample array must be carefully controlled. While sample preparation is important, data acquisition will play a critical role in the experiment. Low dosage techniques could be important, as heat distortion of the lattice and radiation damage will induce severe systematic errors in diffraction data. However, the advantage of X-ray over electron diffraction is that X-rays will transfer less heat to the sample than electrons. The specimen can be scanned to identify regions of high crystallinity. The membranes will then be tilted at various angles so that different views of the structure and additional diffraction data relating to electron density modulations perpendicular to the crystal plane can be obtained. Sample preparation for sub-micron 3-D crystals is being developed at the CSIRO Health Sciences and Nutrition structural biology laboratory that is participating in the Centre. This presents technical challenges with respect to mounting sub-micron crystals in a liquid nitrogen temperature cold stream and aligning them with the proposed X-ray source. Data can be collected from 3-D crystals by using the screenless oscillation method and the imaging methods described below will generate an electron density map of the protein to be interpreted in terms of the protein sequence and any post-translation modifications. The laboratory maintains an extensive UNIX based system for crystallographic software, including graphics workstations and Beowulf clusters.