SIGNAL TRANSDUCTION. Structural basis for nucleotide exchange in heterotrimeric G proteins.

G protein-coupled receptors (GPCRs) relay diverse extracellular signals into cells by catalyzing nucleotide release from heterotrimeric G proteins, but the mechanism underlying this quintessential molecular signaling event has remained unclear. Here we use atomic-level simulations to elucidate the nucleotide-release mechanism. We find that the G protein α subunit Ras and helical domains-previously observed to separate widely upon receptor binding to expose the nucleotide-binding site-separate spontaneously and frequently even in the absence of a receptor. Domain separation is necessary but not sufficient for rapid nucleotide release. Rather, receptors catalyze nucleotide release by favoring an internal structural rearrangement of the Ras domain that weakens its nucleotide affinity. We use double electron-electron resonance spectroscopy and protein engineering to confirm predictions of our computationally determined mechanism.

A pipeline for comprehensive and automated processing of electron diffraction data in IPLT.

Electron crystallography of two-dimensional crystals allows the structural study of membrane proteins in their native environment, the lipid bilayer. Determining the structure of a membrane protein at near-atomic resolution by electron crystallography remains, however, a very labor-intense and time-consuming task. To simplify and accelerate the data processing aspect of electron crystallography, we implemented a pipeline for the processing of electron diffraction data using the Image Processing Library and Toolbox (IPLT), which provides a modular, flexible, integrated, and extendable cross-platform, open-source framework for image processing. The diffraction data processing pipeline is organized as several independent modules implemented in Python. The modules can be accessed either from a graphical user interface or through a command line interface, thus meeting the needs of both novice and expert users. The low-level image processing algorithms are implemented in C++ to achieve optimal processing performance, and their interface is exported to Python using a wrapper. For enhanced performance, the Python processing modules are complemented with a central data managing facility that provides a caching infrastructure. The validity of our data processing algorithms was verified by processing a set of aquaporin-0 diffraction patterns with the IPLT pipeline and comparing the resulting merged data set with that obtained by processing the same diffraction patterns with the classical set of MRC programs.

Simulation and correction of electron images of tilted planar weak-phase samples.

The phase contrast theory describes the transfer of information from a weak-phase object to the image plane of a transmission electron microscope. For a tilted sample where the distance from the focal plane varies continuously across the field of view, the recently introduced Tilted Contrast Imaging Function (TCIF) model provides the mathematical description of this information transfer. Here we expand the TCIF model to account for astigmatism, and present several methods to generate simulated images of tilted samples and compare them to experimental results. We analyze in depth the differences between TCIF and the classical Contrast Transfer Function (CTF) model, which assumes invariant defocus, and discuss how they can affect the interpretation of experimental data. In addition, we apply the TCIF model to simulated test objects in order to explore the performance of techniques that aim to correct the artifacts introduced by the imaging function, and evaluate how well they recover the original information after optimizing the parameters.

OpenStructure: a flexible software framework for computational structural biology.

MOTIVATION: Developers of new methods in computational structural biology are often hampered in their research by incompatible software tools and non-standardized data formats. To address this problem, we have developed OpenStructure as a modular open source platform to provide a powerful, yet flexible general working environment for structural bioinformatics. OpenStructure consists primarily of a set of libraries written in C++ with a cleanly designed application programmer interface. All functionality can be accessed directly in C++ or in a Python layer, meeting both the requirements for high efficiency and ease of use. Powerful selection queries and the notion of entity views to represent these selections greatly facilitate the development and implementation of algorithms on structural data. The modular integration of computational core methods with powerful visualization tools makes OpenStructure an ideal working and development environment. Several applications, such as the latest versions of IPLT and QMean, have been implemented based on OpenStructure-demonstrating its value for the development of next-generation structural biology algorithms. AVAILABILITY: Source code licensed under the GNU lesser general public license and binaries for MacOS X, Linux and Windows are available for download at http://www.openstructure.org. CONTACT: torsten.schwede@unibas.ch SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Projection maps of three members of the KdgM outer membrane protein family.

We present the projection structures of the three outer membrane porins KdgM and KdgN from Erwinia chrysanthemi and NanC from Escherichia coli, based on 2D electron crystallography. A wide screening of 2D crystallization conditions yielded tubular crystals of a suitable size and quality to perform high-resolution electron microscopy. Data processing of untilted samples allowed us to separate the information of the two crystalline layers and resulted in projection maps to a resolution of up to 7A. All three proteins exhibit a similar putative beta-barrel structure and the three crystal forms have the same symmetry. However, there are differences in the packing arrangements of the monomers as well as the densities of the projections. To interpret these projections, secondary structure prediction was performed using beta-barrel specific prediction algorithms. The predicted transmembrane beta-barrels have a high similarity in the arrangement of the putative beta-strands and the loops, but do not match those of OmpG, a related protein porin whose structure was solved.

Function and molecular architecture of the Yersinia injectisome tip complex.

By quantitative immunoblot analyses and scanning transmission electron microscopy (STEM), we determined that the needle of the Yersinia enterocolitica E40 injectisome consists of 139 +/- 19 YscF subunits and that the tip complex is formed by three to five LcrV monomers. A pentamer represented the best fit for an atomic model of this complex. The N-terminal globular domain of LcrV forms the base of the tip complex, while the central globular domain forms the head. Hybrids between LcrV and its orthologues PcrV (Pseudomonas aeruginosa) or AcrV (Aeromonas salmonicida) were engineered and recombinant Y. enterocolitica expressing the different hybrids were tested for their capacity to form the translocation pore by a haemolysis assay. There was a good correlation between haemolysis, insertion of YopB into erythrocyte membranes and interaction between YopB and the N-terminal globular domain of the tip complex subunit. Hence, the base of the tip complex appears to be critical for the functional insertion of YopB into the host cell membrane.

The contrast-imaging function for tilted specimens.

A theoretical description of the contrast-imaging function is derived for tilted specimens that exhibit weak-phase object characteristics. We show that the tilted contrast-imaging function (TCIF) is a linear transformation, which can be approximated by the convolution operation for small tilt angles or for small specimens. This approximation is not valid for electron tomography, where specimen tilts are above 60 degrees and specimen dimensions amount to some 10 microm. The approximation also breaks down for electron crystallography, where atomic resolution is to be achieved. Therefore, we do not make this approximation and propose a generalized algorithm for inverting the TCIF. The implications of our description are discussed in the context of electron tomography, single particle analysis, and electron crystallography, and the improved resolution is quantitatively demonstrated.

Collaborative EM image processing with the IPLT image processing library and toolbox.

We present the Image Processing Library and Toolbox, IPLT, in the context of a collaborative electron microscopy processing effort, which has driven the evolution of our software architecture over the last years. The high-level interface design as well as the underlying implementations are described to demonstrate the flexibility of the IPLT framework. It aims to support the wide range of skills and interests of methodologically oriented scientists who wish to implement their ideas and algorithms as processing code.

Cytotoxin ClyA from Escherichia coli assembles to a 13-meric pore independent of its redox-state.

ClyA is a pore-forming toxin from virulent Escherichia coli and Salmonella enterica strains. Here, we show that the intrinsic hemolytic activity of ClyA is independent of its redox state, and that the assembly of both reduced and oxidized ClyA to the ring-shaped oligomer is triggered by contact with lipid or detergent. A rate-limiting conformational transition in membrane-bound ClyA monomers precedes their assembly to the functional pore. We obtained a three-dimensional model of the detergent-induced oligomeric complex at 12 A resolution by combining cryo- and negative stain electron microscopy with mass measurements by scanning transmission electron microscopy. The model reveals that 13 ClyA monomers assemble into a cylinder with a hydrophobic cap region, which may be critical for membrane insertion.

Structure of the rhodopsin dimer: a working model for G-protein-coupled receptors.

G-protein-coupled receptors (GPCRs) participate in virtually all physiological processes. They constitute the largest and most structurally conserved family of signaling molecules. Several class C GPCRs have been shown to exist as dimers in their active form and growing evidence indicates that many, if not all, class A receptors also form dimers and/or higher-order oligomers. High-resolution crystal structures are available only for the detergent-solubilized light receptor rhodopsin (Rho), the archetypal class A GPCR. In addition, Rho is the only GPCR for which the presumed higher-order oligomeric state has been demonstrated, by imaging native disk membranes using atomic force microscopy (AFM). Based on these data and the X-ray structure, an atomic model of Rho dimers has been proposed, a model that is currently scrutinized in various ways. AFM has also been used to measure the forces required to unfold single Rho molecules, thereby revealing which residues are responsible for Rho's stability. Recent functional analyses of fractions from solubilized disk membranes revealed that higher-order Rho oligomers are the most active species. These and other results have enhanced our understanding of GPCR structure and function.

The 4.5 A structure of human AQP2.

Located in the principal cells of the collecting duct, aquaporin-2 (AQP2) is responsible for the regulated water reabsorption in the kidney and is indispensable for the maintenance of body water balance. Disregulation or malfunctioning of AQP2 can lead to severe diseases such as nephrogenic diabetes insipidus, congestive heart failure, liver cirrhosis and pre-eclampsia. Here we present the crystallization of recombinantly expressed human AQP2 into two-dimensional protein-lipid arrays and their structural characterization by atomic force microscopy and electron crystallography. These crystals are double-layered sheets that have a diameter of up to 30 microm, diffract to 3 A(-1) and are stacked by contacts between their cytosolic surfaces. The structure determined to 4.5 A resolution in the plane of the membrane reveals the typical aquaporin fold but also a particular structure between the stacked layers that is likely to be related to the cytosolic N and C termini.

Structure of zinc-independent sorbitol dehydrogenase from Rhodobacter sphaeroides at 2.4 A resolution.

Recombinant sorbitol dehydrogenase (SDH) from Rhodobacter sphaeroides has been crystallized in the absence of the cofactor NAD(H) and its structure determined to 2.4 A resolution using molecular replacement (refined R and R free factors of 18.8 and 23.8%, respectively). As expected from the sequence and shown by the conserved fold, SDH can be assigned to the short-chain dehydrogenase/reductase protein family. The cofactor NAD and the substrate sorbitol have been modelled into the structure and the active-site architecture, which displays the highly conserved catalytic tetrad of Asn-Ser-Tyr-Lys residues, is discussed in relation to the enzyme mechanism. This is the first structure of a bacterial SDH belonging to the SDR family.

Structure and assembly of the pseudopilin PulG.

The pseudopilin PulG is one of several essential components of the type II pullulanase secretion machinery (the Pul secreton) of the Gram-negative bacterium Klebsiella oxytoca. The sequence of the N-terminal 25 amino acids of the PulG precursor is hydrophobic and very similar to the corresponding region of type IV pilins. The structure of a truncated PulG (lacking the homologous region), as determined by X-ray crystallography, was found to include part of the long N-terminal alpha-helix and the four internal anti-parallel beta-strands that characterize type IV pilins, but PulG lacks the highly variable loop region with a disulphide bond that is found in the latter. When overproduced, PulG forms flexible pili whose structural features, as visualized by electron microscopy, are similar to those of bacterial type IV pili. The average helical repeat comprises 17 PulG subunits and four helical turns. Electron microscopy and molecular modelling show that PulG probably assembles into left-handed helical pili with the long N-terminal alpha-helix tightly packed in the centre of the pilus. As in the type IV pilins, the hydrophobic N-terminal part of the PulG alpha-helix is necessary for its assembly. Subtle sequence variations within this highly conserved segment seem to determine whether or not a type IV pilin can be assembled into pili by the Pul secreton.

Structural analysis of the reaction center light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy.

The bacterium Rhodospirillum rubrum contains a simple photosynthetic system, in which the reaction center (RC) receives energy from the light-harvesting (LH1) complex. We have used high-resolution atomic force microscopy (AFM) to image two-dimensional crystals of the RC-LH1 complex of R. rubrum. The AFM topographs show that the RC-LH1 complex is approximately 94 A in height, the RC-H subunit protrudes from the cytoplasmic face of the membrane by 40 A, and it sits 21 A above the highest point of the surrounding LH1 ring. In contrast, the RC on the periplasmic side is at a lower level than LH1, which protrudes from the membrane by 12 A. The RC-LH1 complex can adopt an irregular shape in regions of uneven packing forces in the crystal; this reflects a likely flexibility in the natural membrane, which might be functionally important by allowing the export of quinol formed as a result of RC photochemistry. Nanodissection of the RC by the AFM tip removes the RC-H subunit and reveals the underlying RC-L and -M subunits. LH1 complexes completely lacking the RC were also found, providing ideal conditions for imaging both rings of LH1 polypeptides for the first time by AFM. In addition, we demonstrate the ellipticity of the LH1 ring at the cytoplasmic and periplasmic sides of the membrane, in both the presence and absence of the RC. These AFM measurements have been reconciled with previous electron microscopy and NMR data to produce a model of the RC-LH1 complex.

Iplt--image processing library and toolkit for the electron microscopy community.

We present the foundation for establishing a modular, collaborative, integrated, open-source architecture for image processing of electron microscopy images, named iplt. It is designed around object oriented paradigms and implemented using the programming languages C++ and Python. In many aspects it deviates from classical image processing approaches. This paper intends to motivate developers within the community to participate in this on-going project. The iplt homepage can be found at http://www.iplt.org.

Assessing the structure of membrane proteins: combining different methods gives the full picture.

The rotor stoichiometry of F-ATPases has been revealed by the combined approaches of X-ray diffraction (XRD), electron crystallography, and atomic force microscopy (AFM). XRD showed the rotor from the yeast mitochondrial F-ATPase to contain 10 subunits. AFM was used to visualize the tetradecameric chloroplast rotors, and electron crystallography and AFM together revealed the rotors from Ilyobacter tartaricus to be composed of 11 subunits. While biochemical methods had determined an approximate stoichiometric value, precise measurements and new insights into a species-dependent rotor stoichiometry became available by applying the three structural tools together. The structures of AQP1, a water channel, and G1pF, a glycerol channel, were determined by electron crystallography and XRD. The combination of both of these structural tools with molecular dynamics simulations gave a differentiated description of the mechanisms determining the selectivity of water and glycerol channels. This illustrates that the combination of different methods in structural biology reveals more than each method alone.

Structure and function of water channels.

All aquaporins are efficient water transporters, while sustaining strict selectivity, even against protons, thereby maintaining the proton gradient across the cell membrane. Recently solved structures of these membrane channels have helped us to understand this remarkable property.

Identification and structure of a putative Ca2+-binding domain at the C terminus of AQP1.

Aquaporin-1 (AQP1) is the first functionally identified aquaporin of a growing family of membrane water channels found in all forms of life. Recently, a possible secondary function as a cyclic guanosine monophosphate (cGMP) gated ion channel was attributed to AQP1. We have reconstituted purified protein from bovine and human red blood cell membranes into highly ordered 2D crystals. The topography of both AQP1s was determined by electron microscopy from freeze-dried, unidirectionally metal-shadowed 2D crystals as well as from surface topographs of native crystals recorded in buffer solution with the atomic force microscope (AFM). In spite of the high level of sequence homology between bovine and human AQP1, the surfaces showed distinct differences. Alignment of both sequences and comparison of the acquired surface topographies with the atomic model of human AQP1 revealed the topographic changes on the surface of bovine AQP1 to be induced by a few amino acid substitutions. A striking degree of sequence homology was found between the carboxyl-terminal domains of AQP1s from different organisms and EF-hands from Ca2+-binding proteins belonging to the calmodulin superfamily, suggesting the existence of a Ca2+-binding site at the C terminus of AQP1 instead of the putative cGMP-binding site reported previously. To unveil its position on the acquired surface topographies, 2D crystals of AQP1 were digested with carboxypeptidase Y, which cleaves off the intracellular C terminus. Difference maps of AFM topographs between the native and the peptidase-treated AQP1s showed the carboxylic tail to be close to the 4-fold symmetry axis of the tetramer. SDS-PAGE and matrix-assisted laser desorption/ionisation mass spectrometry of native and decarboxylated bovine and human AQP1 revealed that the EF-hand motif found at the C terminus of AQP1 was partially resistant to peptidase digestion. The importance of the C-terminal domain is implicated by structural instability of decarboxylated AQP1. A possible role of the C terminus and calcium in translocation of AQP1 in cholangiocytes from intracellular vesicles to the plasma membrane and in triggering its fusion is discussed. Functional studies are now required to identify the physiological role of the Ca2+-binding site.

Sugar transport through maltoporin of Escherichia coli: role of the greasy slide.

The lining of the maltodextrin-specific maltoporin (LamB) channel exhibits a string of aromatic residues, the greasy slide, part of which has been shown previously by crystallography to be involved in substrate binding. To probe the functional role of the greasy slide, alanine scanning mutagenesis has been performed on the six greasy slide residues and Y118 at the channel constriction. The mutants were characterized by an in vivo uptake assay and sugar-induced-current-noise analysis. Crystallographic analysis of the W74A mutant showed no perturbation of the structure. All mutants showed considerably decreased maltose uptake rates in vivo (<10% of the wild-type value), indicating the functional importance of the investigated residues. Substitutions at the channel center revealed appreciably increased (up to 100-fold) in vitro half-saturation concentrations for maltotriose and maltohexaose binding to the channel. Sugar association rates, however, were significantly affected also by the mutations at either end of the slide (W74A, W358A, and F227A), an effect which became most apparent upon nonsymmetrical sugar addition. The kinetic data are discussed on the basis of an asymmetric one-site two-barrier model, which suggests that, at low substrate concentrations, as are found under physiological conditions, only the heights of the extracellular and periplasmic barriers, which are reduced by the presence of the greasy slide, determine the efficiency of this facilitated diffusion channel.

Imaging the electrostatic potential of transmembrane channels: atomic probe microscopy of OmpF porin.

The atomic force microscope (AFM) was used to image native OmpF porin and to detect the electrostatic potential generated by the protein. To this end the OmpF porin trimers from Escherichia coli was reproducibly imaged at a lateral resolution of approximately 0.5 nm and a vertical resolution of approximately 0.1 nm at variable electrolyte concentrations of the buffer solution. At low electrolyte concentrations the charged AFM probe not only contoured structural details of the membrane protein surface but also interacted with local electrostatic potentials. Differences measured between topographs recorded at variable ionic strength allowed mapping of the electrostatic potential of OmpF porin. The potential map acquired by AFM showed qualitative agreement with continuum electrostatic calculations based on the atomic OmpF porin embedded in a lipid bilayer at the same electrolyte concentrations. Numerical simulations of the experimental conditions showed the measurements to be reproduced quantitatively when the AFM probe was included in the calculations. This method opens a novel avenue to determine the electrostatic potential of native protein surfaces at a lateral resolution better than 1 nm and a vertical resolution of approximately 0.1 nm.