Normal-mode refinement of anisotropic thermal parameters for potassium channel KcsA at 3.2 A crystallographic resolution

We report a normal-mode method for anisotropic refinement of membrane-protein structures, based on a hypothesis that the global near-native-state disordering of membrane proteins in crystals follows low-frequency normal modes. Thus, a small set of modes is sufficient to represent the anisotropic thermal motions in X-ray crystallographic refinement. By applying the method to potassium channel KcsA at 3.2 A, we obtained a structural model with an improved fit with the diffraction data. Moreover, the improved electron density maps allowed for large structural adjustments for 12 residues in each subunit, including the rebuilding of 3 missing side chains. Overall, the anisotropic KcsA structure at 3.2 A was systematically closer to a 2.0 A KcsA structure, especially in the selectivity filter. Furthermore, the anisotropic thermal ellipsoids from the refinement revealed functionally relevant structural flexibility. We expect this method to be a valuable tool for structural refinement of many membrane proteins with moderate-resolution diffraction data.     

X-ray diffraction from a single layer of purple membrane at the air/water interface

X-ray diffraction patterns have been recorded from a single layer of purple membrane ( approximately 50 A thickness) at the air/water interface in a Langmuir trough. Grazing-incidence X-ray diffraction is demonstrated to be a promising method for obtaining structural information on membrane proteins under physiological conditions. The method is so sensitive that diffraction can be measured from samples with only 10(13) protein molecules in the beam. Diffraction from hexagonal crystals of purple membrane with a lattice constant of 61. 3 A was observed up to the order {h,k}={4,3}, corresponding to a resolution of approximately 9 A. The work reported here is a first step towards a new way of protein crystallography using grazing-incidence X-ray diffraction at the air/water interface.     

High-resolution membrane protein structure by joint calculations with solid-state NMR and X-ray experimental data

X-ray diffraction and nuclear magnetic resonance spectroscopy (NMR) are the staple methods for revealing atomic structures of proteins. Since crystals of biomolecular assemblies and membrane proteins often diffract weakly and such large systems encroach upon the molecular tumbling limit of solution NMR, new methods are essential to extend structures of such systems to high resolution. Here we present a method that incorporates solid-state NMR restraints alongside of X-ray reflections to the conventional model building and refinement steps of structure calculations. Using the 3.7 Å crystal structure of the integral membrane protein complex DsbB-DsbA as a test case yielded a significantly improved backbone precision of 0.92 Å in the transmembrane region, a 58% enhancement from using X-ray reflections alone. Furthermore, addition of solid-state NMR restraints greatly improved the overall quality of the structure by promoting 22% of DsbB transmembrane residues into the most favored regions of Ramachandran space in comparison to the crystal structure. This method is widely applicable to any protein system where X-ray data are available, and is particularly useful for the study of weakly diffracting crystals.     

Crystallization of the human, mitochondrial voltage-dependent anion-selective channel in the presence of phospholipids.

Overexpressed human voltage-dependent anion-selective channel VDAC or porin from mitochondrial outer membranes has been purified to homogeneity. Electron microscopic analysis of VDAC in detergent solution revealed a uniform particle population consisting of porin monomers. After dialysis of detergent-solubilized porin in the presence of dimyristoylphosphatidylcholine at lipid-to-protein ratios between 0.2 and 0.5 (percentage by weight), mostly multilamellar crystals were obtained. Crystals adsorbed to carbon films flattened during negative staining and air-drying and exhibited different structural features due to differences in the vertical stacking of several crystalline layers, each consisting of one membrane bilayer. Adsorbed, frozen-hydrated multilamellar membrane crystals revealed uniform diffraction patterns with sharp diffraction spots extending to 8.2 A. The surface structure of VDAC was reconstructed from freeze-dried and unidirectionally metal-shadowed crystals. Major protein protrusions were observed from two VDAC monomers present in the unit cell. Differences in the surface structural features indicate alternate orientations of VDAC molecules with respect to the lipid bilayer, allowing the simultaneous imaging of both the cytosolic and intramitochondrial surfaces. Each VDAC molecule consists of a pore lumen with a diameter of 17-20 A surrounded by a protein rim of nonuniform height, suggesting an asymmetrical distribution of protein mass around the diffusion channels.     

Refined structure of the pore-forming domain of colicin A at 2.4 A resolution.

The E1 subgroup (E1, A, B, IA, IB, K and N) of anti-bacterial toxins called colicins is known to form voltage-dependent channels in lipid bilayers. The crystal structure of the pore-forming domain of colicin A from Escherichia coli has been refined to the diffraction limit of the crystals at 2.4 A resolution by means of molecular dynamics and restrained least-squares methods to a conventional R-factor of 0.18 for all data between 6.0 and 2.4 A resolution. The polypeptide chain of 204 amino acid residues consists of ten alpha-helices organized in a three-layer structure. The helices range in length from 9 to 23 residues with an average length of 125 residues. The packing arrangement of the helices has been analysed; the packing is different from that observed in four-helix bundle proteins. The sites of 83 water molecules have been located and refined. Analysis of the structure provides insights into the mechanism of formation of a voltage-gated channel by the protein. Although it is proposed that substantial tertiary structural changes occur during membrane insertion, the secondary structural elements remain conserved. This idea has been proposed recently for a number of other protein-membrane events and thus may have more general applicability.     

Structure of gramicidin A.

Gramicidin A, a hydrophobic linear polypeptide, forms channels in phospholipid membranes that are specific for monovalent cations. Nuclear Magnetic Resonance (NMR) spectroscopy provided the first direct physical evidence that the channel conformation in membranes is an amino terminal-to-amino terminal helical dimer, and circular dichroism (CD) spectroscopy has shown the sensitivity of its conformation to different environments and the structural consequences of ion binding. The three-dimensional structure of a gramicidin/cesium complex has been determined by x-ray diffraction of single crystals using single wavelength anomalous scattering for phasing. The left-handed double helix in this crystal form corresponds to one of the intermediates in the process of folding and insertion into membranes. Co-crystals of gramicidin and lipid that appear to have gramicidin in their membrane channel conformation have also been formed and are presently under investigation. Hence, we have used a combination of spectroscopic and diffraction techniques to examine the conformation and functionally-related structural features of gramicidin A.     

Designed ankyrin repeat protein binders for the crystallization of AcrB: plasticity of the dominant interface

The formation of well-diffracting crystals is a major bottleneck in structural analysis of membrane proteins by X-ray crystallography. One approach to improve crystal quality is the use of DARPins as crystallization chaperones. Here, we present a detailed analysis of the interaction between DARPins and the integral membrane protein AcrB. We find that binders selected in vitro by ribosome display share a common epitope. The comparative analysis of three crystal structures of AcrB-DARPin complexes allowed us to study the plasticity of the interaction with this dominant binding site. Seemingly redundant AcrB-DARPin crystals show substantially different diffraction quality as a result of subtle differences in the binding geometry. This work exemplifies the importance to screen a number of crystallization chaperones to obtain optimal diffraction data. Crystallographic analysis is complemented by biophysical characterization of nine AcrB binders. We observe that small variations in the interface can lead to differing behavior of the DARPins with regards to affinity, stoichiometry of the complexes and specificity for their target.     

A protocol for high throughput methods for the expression and purification of inner membrane proteins

The production of diffraction quality crystals for the structural determination of inner membrane proteins relies on obtaining large amounts of stable protein. Achieving this, by finding the correct parameters to successfully express and purify these proteins is often time-consuming and frustrating. The methods described here examine the most important parameters, in both expression and purification, quickly and simply. They take into account methods previously used in successful structural determinations of inner membrane proteins and collect and analyse data for use in further experiments and to investigate overall trends. These methods make use of histidine-tagged membrane proteins with a green fluorescent protein fusion but could be adapted easily for other proteins.     

Two-dimensional crystallization of the ryanodine receptor Ca2+ release channel on lipid membranes

The ryanodine receptor (RyR) is the largest known membrane protein with a total molecular mass of 2.3 x 10(3) kDa. Well ordered, two-dimensional (2D) crystals are an essential prerequisite to enable RyR structure determination by electron crystallography. Conventionally, the 2D crystallization of membrane proteins is based on a 'trial-and-error' strategy, which is both time-consuming and chance-directed. By adopting a new strategy that utilizes protein sequence information and predicted transmembrane topology, we successfully crystallized the RyR on positively charged lipid membranes. Image processing of negatively stained crystals reveals that they are well ordered, with diffraction spots of IQ < or = 4 extending to approximately 20 angstroms, the resolution attainable in negative stain. The RyR crystals obtained on the charged lipid membrane have characteristics consistent with 2D arrays that have been observed in native sarcoplasmic reticulum of muscle tissues. These crystals provide ideal materials to enable structural analysis of RyR by high-resolution electron crystallography. Moreover, the reconstituted native-like 2D array provides an ideal model system to gain structural insights into the mechanism of RyR-mediated Ca2+ signaling processes, in which the intrinsic ability of RyR oligomers to organize into a 2D array plays a crucial role.     

Fragment-based phase extension for three-dimensional structure determination of membrane proteins by electron crystallography

In electron crystallography, membrane protein structure is determined from two-dimensional crystals where the protein is embedded in a membrane. Once large and well-ordered 2D crystals are grown, one of the bottlenecks in electron crystallography is the collection of image data to directly provide experimental phases to high resolution. Here, we describe an approach to bypass this bottleneck, eliminating the need for high-resolution imaging. We use the strengths of electron crystallography in rapidly obtaining accurate experimental phase information from low-resolution images and accurate high-resolution amplitude information from electron diffraction. The low-resolution experimental phases were used for the placement of α helix fragments and extended to high resolution using phases from the fragments. Phases were further improved by density modifications followed by fragment expansion and structure refinement against the high-resolution diffraction data. Using this approach, structures of three membrane proteins were determined rapidly and accurately to atomic resolution without high-resolution image data.     

X-ray diffraction of heavy-atom labelled two-dimensional crystals of rhodopsin identifies the position of cysteine 140 in helix 3 and cysteine 316 in helix 8

We have used site-specific heavy-atom labelling and X-ray diffraction to localize single amino acid residues in the cytoplasmic domain of the integral membrane protein rhodopsin, the dim-light photoreceptor of retinal vertebrate rod cells. Two-dimensional orthorhombic crystals of the space group p22(1)2(1) (a=59.5(+/-1) A and b=82.7(+/-1.5) A) were produced from detergent-solubilized, partially delipidated rhodopsin. To obtain milligram amounts of two-dimensional crystals, which are required for X-ray diffraction, the yield of the crystalline material was significantly increased by reconstitution of rhodopsin in the presence of cholesterol (1:2 to 1:10 mol/mol) and by adding polar organic solvents to the dialysis buffer. The native cysteine residues C140 and C316 were then selectively labelled with mercury using the sulphydryl-specific reagent p-chloromercuribenzoate (1.6-2.1 mol Hg per mol rhodopsin). The labelling did not affect the unit cell dimensions. Optical absorption spectra of labelled and native two-dimensional rhodopsin crystals showed the characteristic 11-cis-retinal peak at 498 nm, which corresponds to the dark state of rhodopsin. The in-plane position of the mercury label was calculated at 9.5 A resolution from the intensity differences in the X-ray diffraction patterns of labelled and native crystals using Fourier difference methods and the phase information from electron crystallography. The label positions were in excellent agreement with the positions of C140 at the cytoplasmic end of helix 3 and of C316 in the cytoplasmic helix 8 recently obtained from three-dimensional rhodopsin crystals. Whereas these high-resolution diffraction studies were performed under cryogenic conditions (100 K), our results were obtained at room temperature with fully hydrated membranes and in the absence of loop-loop crystal contacts. To study the structural changes of the cytoplasmic loops involved in activation and signal transduction, our more physiological conditions offer important advantages. Furthermore, the localization of C316 is the first direct proof that the electron density on top of helix 1 observed by cryo-electron microscopy is a part of the C-terminal loop. Our approach is of particular interest for investigations of other membrane proteins, for which 3D crystals are not available. Structural constraints from heavy-atom labels at strategic sites enable the assignment of a position in the amino acid sequence to features visible in a low-resolution density map and the study of conformational changes associated with different functional states of the membrane protein.     

Two-dimensional structure of plant light-harvesting complex at 3.7 A [corrected] resolution by electron crystallography

The structure of the light-harvesting chlorophyll a/b-protein complex has been determined at 3.7 A resolution in projection by electron diffraction, electron microscopy and image analysis. Diffraction patterns and high-resolution spotscan images of two-dimensional crystals stabilized with tannin were recorded at low temperature. Phases of structure factors were obtained directly by image processing, after correction of the images for lattice distortions, defocus and beam tilt. Amplitudes were measured by electron diffraction. The projection map shows the detailed structure of the trimeric complex, suggesting the positions of two domains of potential structural and functional homology, of one membrane-spanning alpha-helix approximately perpendicular to the membrane plane and of several tightly bound lipid molecules.     

Room to move: crystallizing membrane proteins in swollen lipidic mesophases

The cubic phase or in meso crystallization method is responsible for almost 40 solved integral membrane protein structures. Most of these are small and compact proteins. A model for how crystals form by the in meso method has been proposed that invokes a transition between mesophases. In light of this model, we speculated that a more hydrated and open mesophase, of reduced interfacial curvature, would support facile crystallization of bigger and bulkier proteins. The proposal was explored here by performing crystallization in the presence of additives that swell the cubic phase. The additive concentration inducing swelling, as quantified by small-angle X-ray diffraction, coincided with a "crystallization window" in which two, very different transmembranal proteins produced crystals. That the swollen mesophase can grow structure-grade crystals was proven with one of these, the light-harvesting II complex. In most regards, the structural details of the corresponding complex resembled those of crystals grown by the conventional vapour diffusion method, with some important differences. In particular, packing density in the in meso-grown crystals was dramatically higher, more akin to that seen with water-soluble proteins, which accounts for their enhanced diffracting power. The layered and close in-plane packing observed has been rationalized in a model for nucleation and crystal growth by the in meso method that involves swollen mesophases. These results present a rational case for including mesophase-swelling additives in screens for in meso crystallogenesis. Their use will contribute to broadening the range of membrane proteins that yield to structure determination.     

X-ray diffraction from membrane protein nanocrystals

Membrane proteins constitute >30% of the proteins in an average cell, and yet the number of currently known structures of unique membrane proteins is <300. To develop new concepts for membrane protein structure determination, we have explored the serial nanocrystallography method, in which fully hydrated protein nanocrystals are delivered to an x-ray beam within a liquid jet at room temperature. As a model system, we have collected x-ray powder diffraction data from the integral membrane protein Photosystem I, which consists of 36 subunits and 381 cofactors. Data were collected from crystals ranging in size from 100 nm to 2 μm. The results demonstrate that there are membrane protein crystals that contain <100 unit cells (200 total molecules) and that 3D crystals of membrane proteins, which contain <200 molecules, may be suitable for structural investigation. Serial nanocrystallography overcomes the problem of x-ray damage, which is currently one of the major limitations for x-ray structure determination of small crystals. By combining serial nanocrystallography with x-ray free-electron laser sources in the future, it may be possible to produce molecular-resolution electron-density maps using membrane protein crystals that contain only a few hundred or thousand unit cells.     

Densely packed beta-structure at the protein-lipid interface of porin is revealed by high-resolution cryo-electron microscopy

Porin is an integral membrane protein that forms channels across the outer membrane of Escherichia coli. Electron microscopic studies of negatively stained two-dimensional porin crystals have shown three stain accumulations per porin trimer, revealing the locations of pores spanning the membrane. In this study, reconstituted porin lattices embedded in glucose were investigated using the low-dose technique on a cryo-electron microscope equipped with a helium-cooled superconducting objective lens. The specimen temperature was maintained at 5 K to yield an improved microscopic and specimen stability. Under these conditions, we obtained for the first time electron diffraction patterns from porin lattices to a resolution of 3.2 A and images showing optical diffraction up to a resolution of 4.9 A. Applying correlation averaging techniques to the digitized micrographs, we were able to reconstruct projected images of the porin trimer to a resolution of up to 3.5 A. In the final projection maps, amplitudes from electron diffraction and phases from these images were combined. The predominant feature is a high-density narrow band (about 6 A in thickness) that delineates the outer perimeter of the trimer. Since the molecule consists of almost exclusively beta-sheet structure, as revealed by spectroscopic data, we conclude that this band is a cylindrical beta-pleated sheet crossing the membrane nearly perpendicularly to its plane. Another intriguing finding is a low-density area (about 70 A2) situated in the centre of the trimer.     

Lessons from crystals grown in the Advanced Protein Crystallisation Facility for conventional crystallisation applied to structural biology

The crystallographic quality of protein crystals that were grown in microgravity has been compared to that of crystals that were grown in parallel on earth gravity under otherwise identical conditions. A goal of this comparison was to assess if a more accurate 3D-structure can be derived from crystallographic analysis of the former crystals. Therefore, the properties of crystals prepared with the Advanced Protein Crystallisation Facility (APCF) on earth and in orbit during the last decade were evaluated. A statistical analysis reveals that about half of the crystals produced under microgravity had a superior X-ray diffraction limit with respect of terrestrial controls. Eleven protein structures could be determined at previously unachieved resolutions using crystals obtained in the APCF. Microgravity induced features of the most relevant structures are reported. A second goal of this study was to identify the cause of the crystal quality enhancement useful for structure determination. No correlations between the effect of microgravity and other system-dependent parameters, such as isoelectric point or crystal solvent content, were found except the reduced convection during the crystallisation process. Thus, crystal growth under diffusive regime appears to be the key parameter explaining the beneficial effect of microgravity on crystal quality. The mimicry of these effects on earth in gels or in capillary tubes is discussed and the practical consequences for structural biology highlighted.     

Crystals of crotoxin suitable for high resolution x-ray diffraction analysis

The phospholipasic presynaptic neurotoxin, crotoxin, has been crystallized in a morphology suitable for single crystal x-ray diffraction analysis. The conditions for growth and the unit cell parameters (P4(1)22 or P4(3)22, a = b = 38.5 A, c = 256.9 A, 1 molecule/asymmetric unit) are similar to the very thin plate-like crystals which have been studied with electron diffraction and electron microscopy by Chiu and his colleagues (Jeng, T.-W., Chiu, W., Zemlin, F., and Zeitler, E. (1984) J. Mol. Biol. 175, 93 - 97). These two macroscopic crystal morphologies of what is likely to be a very similar, if not identical, lattice structure will permit the complementary application of electron diffraction/microscopy and x-ray diffraction to understanding the structural basis of the interactions between a phospholipasic neurotoxin and its membrane target.     

Electron diffraction analysis of structural changes in the photocycle of bacteriorhodopsin.

Structural changes are central to the mechanism of light-driven proton transport by bacteriorhodopsin, a seven-helix membrane protein. The main intermediate formed upon light absorption is M, which occurs between the proton release and uptake steps of the photocycle. To investigate the structure of the M intermediate, we have carried out electron diffraction studies with two-dimensional crystals of wild-type bacteriorhodopsin and the Asp96-->Gly mutant. The M intermediate was trapped by rapidly freezing the crystals in liquid ethane following illumination with a xenon flash lamp at 5 and 25 degrees C. Here, we present 3.5 A resolution Fourier projection maps of the differences between the M intermediate and the ground state of bacteriorhodopsin. The most prominent structural changes are observed in the vicinity of helices F and G and are localized to the cytoplasmic half of the membrane.     

Nuclear magnetic resonance in the era of structural genomics

Current interests in structural genomics, and the associated need for high through-put structure determination methods, offer an opportunity to examine new nuclear magnetic resonance (NMR) methodology and the impact this methodology can have on structure determination of proteins. The time required for structure determination by traditional NMR methods is currently long, but improved hardware, automation of analysis, and new sources of data such as residual dipolar couplings promise to change this. Greatly improved efficiency, coupled with an ability to characterize proteins that may not produce crystals suitable for investigation by X-ray diffraction, suggests that NMR will play an important role in structural genomics programs.