Electron cryo-microscopic analysis of crystalline cytochrome oxidase

Sparse and incomplete factorial matrices to screen membrane protein 2D crystallization

Electron crystallography is well suited for studying the structure of membrane proteins in their native lipid bilayer environment. This technique relies on electron cryomicroscopy of two-dimensional (2D) crystals, grown generally by reconstitution of purified membrane proteins into proteoliposomes under conditions favoring the formation of well-ordered lattices. Growing these crystals presents one of the major hurdles in the application of this technique. To identify conditions favoring crystallization a wide range of factors that can lead to a vast matrix of possible reagent combinations must be screened. However, in 2D crystallization these factors have traditionally been surveyed in a relatively limited fashion. To address this problem we carried out a detailed analysis of published 2D crystallization conditions for 12 β-barrel and 138 α-helical membrane proteins. From this analysis we identified the most successful conditions and applied them in the design of new sparse and incomplete factorial matrices to screen membrane protein 2D crystallization. Using these matrices we have run 19 crystallization screens for 16 different membrane proteins totaling over 1300 individual crystallization conditions. Six membrane proteins have yielded diffracting 2D crystals suitable for structure determination, indicating that these new matrices show promise to accelerate the success rate of membrane protein 2D crystallization.

An automated pipeline to screen membrane protein 2D crystallization

Electron crystallography relies on electron cryomicroscopy of two-dimensional (2D) crystals and is particularly well suited for studying the structure of membrane proteins in their native lipid bilayer environment. To obtain 2D crystals from purified membrane proteins, the detergent in a protein-lipid-detergent ternary mixture must be removed, generally by dialysis, under conditions favoring reconstitution into proteoliposomes and formation of well-ordered lattices. To identify these conditions a wide range of parameters such as pH, lipid composition, lipid-to-protein ratio, ionic strength and ligands must be screened in a procedure involving four steps: crystallization, specimen preparation for electron microscopy, image acquisition, and evaluation. Traditionally, these steps have been carried out manually and, as a result, the scope of 2D crystallization trials has been limited. We have therefore developed an automated pipeline to screen the formation of 2D crystals. We employed a 96-well dialysis block for reconstitution of the target protein over a wide range of conditions designed to promote crystallization. A 96-position magnetic platform and a liquid handling robot were used to prepare negatively stained specimens in parallel. Robotic grid insertion into the electron microscope and computerized image acquisition ensures rapid evaluation of the crystallization screen. To date, 38 2D crystallization screens have been conducted for 15 different membrane proteins, totaling over 3000 individual crystallization experiments. Three of these proteins have yielded diffracting 2D crystals. Our automated pipeline outperforms traditional 2D crystallization methods in terms of throughput and reproducibility.

Cytochrome c oxidase: chemistry of a molecular machine

The plethora of proposed chemical models attempting to explain the proton pumping reactions catalyzed by the CcO complex, especially the number of recent models, makes it clear that the problem is far from solved. Although we have not discussed all of the models proposed to date, we have described some of the more detailed models in order to illustrate the theoretical concepts introduced at the beginning of this section on proton pumping as well as to illustrate the rich possibilities available for effecting proton pumping. It is clear that proton pumping is effected by conformational changes induced by oxidation/reduction of the various redox centers in the CcO complex. It is for this reason that the CcO complex is called a redox-linked proton pump. The conformational changes of the proton pump cycle are usually envisioned to be some sort of ligand-exchange reaction arising from unstable geometries upon oxidation/reduction of the various redox centers. However, simple geometrical rearrangements, as in the Babcock and Mitchell models are also possible. In any model, however, hydrogen bonds must be broken and reformed due to conformational changes that result from oxidation/reduction of the linkage site during enzyme turnover. Perhaps the most important point emphasized in this discussion, however, is the fact that proton pumping is a directed process and it is electron and proton gating mechanisms that drive the proton pump cycle in the forward direction. Since many of the models discussed above lack effective electron and/or proton gating, it is clear that the major difficulty in developing a viable chemical model is not formulating a cyclic set of protein conformational changes effecting proton pumping (redox linkage) but rather constructing the model with a set of physical constraints so that the proposed cycle proceeds efficiently as postulated. In our discussion of these models, we have not been too concerned about which electron of the catalytic cycle was entering the site of linkage, but merely whether an ET to the binuclear center played a role. However, redox linkage only occurs if ET to the activated binuclear center is coupled to the proton pump. Since all of the models of proton pumping presented here, with the exception of the Rousseau expanded model and the Wikström model, have a maximum stoichiometry of 1 H+/e-, they inadequately explain the 2 H+/e- ratio for the third and fourth electrons of the dioxygen reduction cycle (see Section V.B). One way of interpreting this shortfall of protons is that the remaining protons are pumped by an as yet undefined indirectly coupled mechanism. In this scenario, the site of linkage could be coupled to the pumping of one proton in a direct fashion and one proton in an indirect fashion for a given electron. For a long time, it was assumed that at least some elements of such an indirect mechanism reside in subunit III. While recent evidence argues against the involvement of subunit III in the proton pump, subunit III may still participate in a regulatory and/or structural capacity (Section II.E). Attention has now focused on subunits I and II in the search for residues intimately involved in the proton pump mechanism and/or as part of a proton channel. In particular, the role of some of the highly conserved residues of helix VIII of subunit I are currently being studied by site directed mutagenesis. In our opinion, any model that invokes heme alpha 3 or CuB as the site of linkage must propose a very effective means by which the presumedly fast uncoupling ET to the dioxygen intermediates is prevented. It is difficult to imagine that ET over the short distance from heme alpha 3 or CuB to the dioxygen intermediate requires more than 1 ns. In addition, we expect the conformational changes of the proton pump to require much more than 1 ns (see Section V.B).

Orientation of LamB signal peptides in bilayers: influence of lipid probes on peptide binding and interpretation of fluorescence quenching data

The orientation in lipid bilayers of the signal sequence of the bacterial protein LamB was studied using binding, circular dichroism, and fluorescence quenching experiments. Measurements were made of binding modifications caused by the incorporation of lipid probes (brominated or nitroxide-labeled phospholipids) used in the parallax fluorescence quenching method of determining peptide penetration depth [Abrams, F. S., and London, E. (1992) Biochemistry 31, 5312-5322]. The signal peptide bound to a similar extent to neutral bilayers composed of either egg phosphatidylcholine (EPC) or phosphatidylcholines brominated at various positions on their acyl chains. The fluorescence of a tryptophan in either the 18 or 24 position of the peptide was quenched more by bromines in the 6 and 7 than in the 9 and 10 positions on the lipid hydrocarbon chain. Parallax calculations showed that tryptophan-18 was located only 4 A from the hydrocarbon-water interface, consistent with the peptide adopting a "hammock" configuration in the bilayer, with both termini exposed to the aqueous phase and the central alpha-helix located near the hydrocarbon-water interface. In contrast, the incorporation of 10% nitroxide-labeled lipids into EPC bilayers modified peptide binding in a manner dependent on the position of the nitroxide on the hydrocarbon chain; 7-Doxyl PC reduced the percent peptide bound by about one-half, whereas 12-Doxyl PC had little effect on binding. These binding differences modified tryptophan quenching by these probes, making parallax analysis problematical. In the presence of the positively charged LamB peptide, the incorporation of negatively charged phospholipids into EPC bilayers increased the level of peptide binding and modified tryptophan quenching by nitroxide probes. These results suggest that the nitroxide probe could be partially excluded from negatively charged lipid domains where the peptide preferentially bound. Quite different binding and quenching results were obtained with a negatively charged peptide analogue, showing that the charge on both the peptide and bilayer affects peptide-nitroxide probe interactions.

Non-helical perturbations of the flagellar filament: Salmonella typhimurium SJW117 at 9.6 A resolution

Using a liquid-helium-cooled superconducting electron cryo-microscope, we obtained low-dose images of negatively stained preparations at 4 K and collected structural data to 1/9.6 -1 for flagellar filaments from the strain SJW117 of Salmonella typhimurium (serotype gt). The subunits of this left-handed, straight filament are non-helically perturbed in a pairwise manner. The perturbation corresponds to an alternating conformation in every other row of subunits. These are the 5-start rows and, necessarily, the resulting structure has a seam. The perturbation is not confined to the outside but extends into the structure. We separated the non-symmetric and symmetric parts of the structural data and generated a three-dimensional reconstruction from the latter. The resulting density map is a structure similar in domain organization to the left-handed filament of S. typhimurium SJW1660. Filtered images generated from the non-symmetric component show an ordered and polar structure. The nature of the perturbation was analyzed by model building using a sphere to represent the subunit at low resolution. A lateral shift of approximately 10 degrees mimics the perturbation.

Vectorially oriented monolayers of the cytochrome c/cytochrome oxidase bimolecular complex

Vectorially oriented monolayers of yeast cytochrome c and its bimolecular complex with bovine heart cytochrome c oxidase have been formed by self-assembly from solution. Both quartz and Ge/Si multilayer substrates were chemical vapor deposited with an amine-terminated alkylsiloxane monolayer that was then reacted with a hetero-bifunctional cross-linking reagent, and the resulting maleimide endgroup surface then provided for covalent interactions with the naturally occurring single surface cysteine 102 of the yeast cytochrome c. The bimolecular complex was formed by further incubating these cytochrome c monolayers in detergent-solubilized cytochrome oxidase. The sequential formation of such monolayers and the vectorially oriented nature of the cytochrome oxidase was studied via meridional x-ray diffraction, which directly provided electron density profiles of the protein(s) along the axis normal to the substrate plane. The nature of these profiles is consistent with previous work performed on vectorially oriented monolayers of either cytochrome c or cytochrome oxidase alone. Furthermore, optical spectroscopy has indicated that the rate of binding of cytochrome oxidase to the cytochrome c monolayer is an order of magnitude faster than the binding of cytochrome oxidase to an amine-terminated surface that was meant to mimic the ring of lysine residues around the heme edge of cytochrome c, which are known to be involved in the binding of this protein to cytochrome oxidase.

Projection structure of the cytochrome bo ubiquinol oxidase from Escherichia coli at 6 A resolution

The haem-copper cytochrome oxidases are terminal catalysts of the respiratory chains in aerobic organisms. These integral membrane protein complexes catalyse the reduction of molecular oxygen to water and utilize the free energy of this reaction to generate a transmembrane proton gradient. Quinol oxidase complexes such as the Escherichia coli cytochrome bo belong to this superfamily. To elucidate the similarities as well as differences between ubiquinol and cytochrome c oxidases, we have analysed two-dimensional crystals of cytochrome bo by cryo-electron microscopy. The crystals diffract beyond 5 A. A projection map was calculated to a resolution of 6 A. All four subunits can be identified and single alpha-helices are resolved within the density for the protein complex. The comparison with the three-dimensional structure of cytochrome c oxidase shows the clear structural similarity within the common functional core surrounding the metal-binding sites in subunit I. It also indicates subtle differences which are due to the distinct subunit composition. This study can be extended to a three-dimensional structure analysis of the quinol oxidase complex by electron image processing of tilted crystals.

Three-dimensional structure of the gap junction connexon

The gap junction membrane channel is composed of macular aggregations of intercellular channels permitting the direct intercellular transfer of ions and small molecules. Each intercellular channel is formed by the apposition of two hexameric transmembrane channels (connexons), one from each cell. The interlocking of the two channels occurs extracellularly in a narrow 2.5-nm "gap" separating the junctional membranes. The channel-channel interaction is known to be selective between members of the family of proteins, called connexins, which oligomerize into the connexons. In addition to selectivity, the molecular interfaces involved in the extracellular interactions between connexons must be very congruent, since the intercellular channel must provide high resistances to the leakage of small ions between the channel lumen and the extracellular space. By using a recently developed biochemical procedure for obtaining ordered arrays of connexons from gap junctions split in the extracellular gap, (Ghoshroy, S., D. A. Goodenough, and G. E. Sosinsky. 1994. Preparation, characterization, and structure of half gap junctional layers spit with urea and EGTA. J. Membr. Biol. 146:15-28) a three-dimensional reconstruction of a connexon has been obtained by electron crystallographic methods. This reconstruction emphasizes the structural asymmetry between the extracellular and cytoplasmic domains and assigns lobed structural features to the extracellular domains of the connexon. The implication of our hemichannel structure is discussed in relation to the in vivo state of unpaired connexons, which have been shown to exist in the plasma membrane.

Purification and two-dimensional crystallization of bacterial cytochrome oxidases

A novel strategy which employes chromatography on an immobilized metal ion has been developed for the purification of bacterial cytochrome c and quinol oxidases. Many bacterial oxidase complexes appear to have a natural affinity to bind to the chelated copper ion. A combination of three different chromatographic principles (anion exchange, metal-affinity and gel filtration) makes an effective tool chest for the preparation of homogeneous and protein-chemically pure bacterial oxidases. These preparations have been used for two-dimensional crystallization. Until now, crystals have been obtained using the Paracococcus denitrificans and Rhodobacter sphaeroides cytochrome aa3 and the Escherichia coli cytochrome bo. The crystals diffract to approximately 2.5 nm in negative stain and have potential for further structural studies.

Reversed metal replicas of freeze-dried proteins to be visualized with the scanning tunneling microscope

Scanning tunneling microscopy of metal-coated specimens has become a reliable technique that permits direct three-dimensional visualization of structural details at a level at which individual subunits in protein complexes or even single domains of proteins can be resolved. We describe in this paper a variation of the freeze-drying metal coating procedure that allows us to image with the STM the inner side of the metal replica, previously in contact with the protein molecules. We have tested this new approach with two different well characterized protein systems: freeze-dried two-dimensional crystals of bacteriophage phi 29 connector and the vesicle form of two-dimensional crystals of cytochrome oxidase from beef heart mitochondria. The images obtained have very good contrast and provide direct topographic information of the crystal surface, complementing structural information obtained previously with transmission electron microscopy. The resolution limit is imposed by the size (2-3 nm diameter) and corrugation of the metal grains used to prepare the replica and by the randomness of the metal shadowing.

Forces and factors that contribute to the structural stability of membrane proteins

While a considerable amount of literature deals with the structural energetics of water-soluble proteins, relatively little is known about the forces that determine the stability of membrane proteins. Similarly, only a few membrane protein structures are known at atomic resolution, although new structures have recently been described. In this article, we review the current knowledge about the structural features of membrane proteins. We then proceed to summarize the existing literature regarding the thermal stability of bacteriorhodopsin, cytochrome-c oxidase, the band 3 protein, Photosystem II and porins. We conclude that a fundamental difference between soluble and membrane proteins is the high thermal stability of intrabilayer secondary structure elements in membrane proteins. This property manifests itself as incomplete unfolding, and is reflected in the observed low enthalpies of denaturation of most membrane proteins. By contrast, the extramembranous parts of membrane proteins may behave much like soluble proteins. A brief general account of thermodynamics factors that contribute to the stability of water soluble and membrane proteins is presented.

Bacteriorhodopsin as a model for proton pumps

According to a long-standing hypothesis, membrane pumps function by flip-flopping between two protein conformations that allow alternative access of the ion binding site to the two membrane surfaces. Site-specific mutagenesis, time-resolved spectroscopy and X-ray diffraction confirm this mechanism for bacteriorhodopsin, and implicate change of electrostatic interaction at the active site as the trigger for the global protein conformation change during the proton transport cycle.

Determination of the structures of respiratory enzyme complexes from mammalian mitochondria

Knowledge of the atomic resolution structures of the respiratory enzymes from mammalian mitochondria is likely to be essential for understanding the molecular basis of human diseases involving their dysfunction. Because they are membrane-bound multisubunit assemblies, the determination of their high resolution structures is a rather challenging undertaking. The greatest progress has been made with the ATP synthase. The structure of its catalytic domain, F1-ATPase, has been solved by X-ray crystallography at 2.8 A resolution. It supports a binding change mechanism of catalysis in intact ATP synthase in which the catalytic subunits are in different states of the catalytic cycle at any instant. Interconversion of the states may be achieved by rotation of an alpha-helical domain of the gamma-subunit relative to the alpha 3 beta 3 sub-assembly. The membrane domain of ATP synthase has been purified, and the stalk linking the catalytic and membrane domains has been reassembled in vitro from its constituent subunits. Considerable progress has also been made in analyzing the structure of bovine complex I. It has about 43 different subunits and 42 of them have been sequenced. All of the known prosthetic groups have been localized in its extrinsic membrane arm, which has been split away from the membrane subunits and purified. The next stage is to crystallize the domain and to solve its structure.

Structure and function of a molecular machine: cytochrome c oxidase

Cytochrome c is responsible for over 90% of the dioxygen consumption in the living cell and contributes to the build-up of a proton electrochemical gradient derived by the vectorial transfer of electrons between cytochrome c and molecular oxygen. The metal ions found in cytochrome oxidases play a crucial role in these processes and have been extensively studied. In this review we present and discuss some of the relevant spectroscopic and kinetic properties of the prosthetic groups of cytochrome c oxidase.

Forces and factors that contribute to the structural stability of membrane proteins

While a considerable amount of literature deals with the structural energetics of water-soluble proteins, relatively little is known about the forces that determine the stability of membrane proteins. Similarly, only a few membrane protein structures are known at atomic resolution, although new structures have recently been described. In this article, we review the current knowledge about the structural features of membrane proteins. We then proceed to summarize the existing literature regarding the thermal stability of bacteriorhodopsin, cytochrome-c oxidase, the band 3 protein, Photosystem II and porins. We conclude that a fundamental difference between soluble and membrane proteins is the high thermal stability of intrabilayer secondary structure elements in membrane proteins. This property manifests itself as incomplete unfolding, and is reflected in the observed low enthalpies of denaturation of most membrane proteins. By contrast, the extramembranous parts of membrane proteins may behave much like soluble proteins. A brief general account of thermodynamics factors that contribute to the stability of water soluble and membrane proteins is presented.

Electron transfer and proton pumping in cytochrome oxidase

This article presents an outlook on the structure and function of terminal oxidases, the respiratory enzymes which catalyze the reduction of dioxygen to water in aerobic organisms. The structure of the redox active metals, their interactions with the protein matrix, and their role in electron transfer ligand binding and proton pumping are briefly reviewed.

Projection structure of the CHIP28 water channel in lipid bilayer membranes at 12-A resolution

Osmotic water transport across plasma membranes in erythrocytes and several epithelial cell types is facilitated by CHIP28, a water-selective membrane channel protein. In order to examine the structure of CHIP28 in membranes, large (1.5-2.5-microns diameter), highly ordered, two-dimensional (2-D) crystals of purified and deglycosylated erythrocyte CHIP28 were generated by reconstitution of detergent-solubilized protein into synthetic lipid bilayers via detergent dialysis. Fourier transforms computed from low-dose electron micrographs of such crystals preserved in negative stain display order to 12-A resolution. The crystal lattice is tetragonal (a = b = 99.2 +/- 1.4 A) with plane group symmetry p4g. A projection density map at 12-A resolution defines the molecular boundary and organization of the CHIP28 monomers in the membrane plane. The unit cell contains four CHIP28 dimers, each composed of two oblong-shaped (37 x 25 A ) monomers with opposite orientations. The CHIP28 monomers associate to form tetrameric structures around the 4-fold axes normal to the membrane plane where stain is excluded. The 2-D crystals of CHIP28 display order extending beyond the limit typically achieved by negative staining and therefore may be amenable to high-resolution structure analysis by cryo-electron microscopy.

Cytochrome c oxidase: structural studies by electron microscopy of two-dimensional crystals

Cytochrome c oxidase is a complex integral membrane protein consisting of 13 different polypeptide chains and four metal centers having a total molecular weight of approximately 200,000 daltons. It can be isolated in two 2-dimensional crystalline forms differing in aggregation state of the enzyme. One crystal form consists of cytochrome oxidase dimers (approximately 400,000 daltons) embedded unidirectionally in the lipid bilayer of a collapsed vesicle while the other form consists of crystalline sheets of cytochrome oxidase monomers. Both crystal forms have been studied by electron microscopy during the past two decades, and this paper summarizes the results of early structural studies as well as more recent results applying techniques of cryoelectron microscopy and digital image processing. The structure of frozen-hydrated cytochrome oxidase dimers at 20 A resolution is discussed as well as the packing of monomers within dimers and the site of cytochrome c binding.

Asymmetry and structural changes in ECF1 examined by cryoelectronmicroscopy

The Escherichia coli ATPase (ECF1) has been studied by cryoelectronmicroscopy and an intrinsic asymmetry of the molecule in the hexagonal projection identified. The three beta subunits could be distinguished. One, which we have called beta 1, has a greater density in projection than the other two; the second, beta 2, is of intermediate density in projection, while the third, beta 3, is smeared out in density. These different features of the beta subunits were used to orient images, and the positions of the gamma and epsilon subunits then established. The location of the gamma subunit, as monitored by the central mass, was not fixed. This subunit could be found in positions that followed an arc from close to beta 2 to close to beta 3, a shift of around 10A, with respect to the center of the mass. The location of the epsilon subunit was monitored after reconstituting a complex of epsilon subunit-depleted ECF1 with a mutant epsilon subunit in which His at residue 38 had been replaced by Cys, and this Cys labeled with an approximately 14A gold particle. The epsilon subunit was found in positions described by an arc between an alpha subunit (alpha 1) and the neighboring beta subunit (beta 1), a shift of around 20A, with respect to the center of the gold particle. A nucleotide dependence of the position of the gamma subunit has been established by Gogol, E.P., Johnston, E., Aggeler, R. and Capaldi, R.A. (1990) Proc. Natl. Acad. Sci. USA 87, 9585-9589. A nucleotide dependence of the position of the epsilon subunit is shown here.(ABSTRACT TRUNCATED AT 250 WORDS)

Electron microscopy of cytochrome c oxidase-containing proteoliposomes: imaging analysis of protein orientation and monomer-dimer behaviour

1. Cytochrome c oxidase-containing vesicles were prepared by cholate dialysis using bovine heart cytochrome c oxidase with egg and dioleoylphosphatidylcholine/dioleoylphosphatidylethanolamines (1:1, w/w) at two ratios of phospholipid to protein (25 mg/mg and 10 mg/mg). With each mixture, one or two (FII, FIII) fractions with mostly outward-facing cytochrome aa3 were separated from a fraction (FI) containing mostly inward-facing enzyme and protein-free liposomes by DEAE-Sephacel chromatography. 2. FII and FIII fractions from egg phospholipid mixtures had 60-80% outward-facing enzyme; FII and FIII fractions from dioleoyl phospholipids showed 50-70% outward-facing enzyme. Egg and dioleoyl phospholipid mixtures maintained good respiratory control ratios (8-13) only at the higher lipid/protein ratios. 3. Platinum/carbon replicas of freeze-fractured vesicle surfaces were subjected to image analysis. The results showed two types of membrane projection with average heights of 7.5 nm and 3.5 nm from the fracture plane. The former were more numerous on the convex faces. Calculated areas of the projections indicated the probable presence of both enzyme dimers and higher aggregates. Oxidase dimers may have membrane areas of 70-80 nm2 at the high (7.5 nm) side and 40-50 nm2 on the low (3.5 nm) side. 4. Proteoliposomes prepared with enzyme depleted of subunit III contained predominantly much smaller projecting areas. These probably represent monomers with high side areas of 35-40 nm2 and low side areas of 20-25 nm2. Electron microscopy thus directly confirms the predicted change of aggregation state resulting from subunit depletion. 5. The results are compared with those from two-dimensional crystals. Assuming that the high and low projections are two sides of one family of transmembrane molecules, a total length of 11 nm matches 11-12 nm lengths obtained by crystallography. Our membrane areas match the areas obtained in earlier 'crystal' studies better than the small areas obtained recently by electron cryomicroscopy.

Binuclear centre structure of terminal protonmotive oxidases

The recent proliferation of data obtained from mutant forms of cytochrome oxidase and analogous enzymes has necessitated a re-examination of existing structural models. A new model is proposed, consistent with these data, which brings several protonatable residues (Y244, D298, D300, T309, T316, K319, T326) into the vicinity of the binuclear centre, suggestive of a proton-transferring function. In addition, we also consider those residues which may participate in electron transport between CuA and haem a. We suggest several potential lines of investigation.

Has negative staining still a place in biomacromolecular electron microscopy?

Transmission electron microscopy of proteins has provided molecular- and in a few cases near-atomic-resolution structural information. In this review, we critically evaluate the potential and the limitations in obtaining molecular resolution, particularly with negatively stained specimens, and put these into perspective with cryomicroscopy of unstained frozen-hydrated and sugar-embedded preparations.

Three-dimensional reconstruction of single particles negatively stained or in vitreous ice

The random-conical reconstruction method has been highly successful in three-dimensional imaging of macromolecules under low-dose conditions. This article summarizes the different steps of this technique as applied to molecules prepared with negative staining or vitreous ice, and sketches out the current directions of development. We anticipate that by using new instrumental developments, transfer function correction and computational refinement techniques, a resolution in the range of 7-10 A could ultimately be achieved.

The thickness determination of organic crystals under low dose conditions using electron energy loss spectroscopy

In the 3-dimensional (3-D) reconstruction of protein crystals with variable thicknesses the electron images and diffraction patterns can only be merged if the crystal thickness is known. Measurement of the thickness using the ratio of the number of inelastically scattered electrons to the number of electrons in the zero loss peak can be accomplished with parallel electron energy loss spectrometry (PEELS). A theoretical analysis of the accuracy of the technique on paraffin crystals of different thicknesses is presented. Our experimental studies with paraffin crystals show the feasibility of measuring a single layer of 47A with good accuracy under low dose and low temperature conditions. A simple experimental apparatus is proposed to obtain thicknesses from small regions of unstained protein crystals prior to collecting the 3-D data sets from the unexposed area of the same crystal.

Cloning, sequencing and deletion from the chromosome of the gene encoding subunit I of the aa3-type cytochrome c oxidase of Rhodobacter sphaeroides

The ctaD gene encoding subunit I of the aa3-type cytochrome c oxidase from Rhodobacter sphaeroides has been cloned. The gene encodes a polypeptide of 565 residues which is highly homologous to the sequences of subunit I from other prokaryotic and eukaryotic sources, e.g. 51% identity with that from bovine, and 75% identity with that from Paracoccus denitrificans. The ctaD gene was deleted from the chromosome of R. sphaeroides, resulting in a strain that spectroscopically lacks cytochrome a. This strain maintains about 50% of the cytochrome c oxidase activity of the wild-type strain owing to the presence of an alternate o-type cytochrome c oxidase. The aa3-type oxidase was restored by complementing the chromosomal deletion with a plasmid-borne copy of the ctaD gene. This system is well suited for site-directed mutagenesis probing of the structure and function of cytochrome c oxidase.

Two-dimensional crystallization of membrane proteins

In spite of several great breakthroughs, the overall rate of progress in determining high-resolution structures of membrane proteins has been slow. This is entirely due to the scarcity of suitable, well-ordered crystals. Most membrane proteins are multimeric complexes with a composite molecular mass in excess of 50000 Da which puts them outside the range of current solution NMR techniques. For the foreseeable future, detailed information about the structure of large membrane proteins will therefore depend on crystallographic methods.

The osmochemistry of electron-transfer complexes

Detailed molecular mechanisms of electron transfer-driven translocation of ions and of the generation of electric fields across biological membranes are beginning to emerge. The ideas inherent in the early formulations of the chemiosmotic hypothesis have provided the framework for this understanding and have also been seminal in promoting many of the experimental approaches which have been successfully used. This article is an attempt to review present understanding of the structures and mechanisms of several osmoenzymes of central importance and to identify and define the underlying features which might be of general relevance to the study of chemiosmotic devices.

Electron microscopic analysis of the peripheral and membrane parts of mitochondrial NADH dehydrogenase (complex I)

Two related forms of the respiratory chain NADH dehydrogenase (NADH:ubiquinone reductase or complex I) are synthesized in the mitochondria of Neurospora crassa. Normally growing cells make a large form that consists of 25 subunits encoded by nuclear DNA and six to seven subunits encoded by mitochondrial DNA. Cells grown in the presence of chloramphenicol, however, make a smaller form comprising only 13 subunits, all encoded by nuclear DNA. When the large enzyme is dissected by chaotropic agents (such as NaBr), all those subunits of the large form that are missing in the small form can be isolated as a distinct, so-called hydrophobic fragment. The small enzyme and the hydrophobic fragment make up, with regard to their redox groups, subunit composition and function, two complementary parts of the large-form NADH dehydrogenase. Averaging of electron microscope images of single particles of the large enzyme was carried out, revealing an unusual L-shaped structure with two domains or "arms" arranged at right angles. The hydrophobic fragment obtained by the NaBr treatment corresponds in size and appearance to one of these arms. A three-dimensional reconstruction from images of negatively stained membrane crystals of the large-form NADH dehydrogenase shows a peripheral domain, protruding from the membrane, with weak unresolved density within the membrane. This peripheral domain was removed by washing the crystals in situ with 2 M-NaBr, exposing a large membrane-buried domain, which was reconstructed in three dimensions. A three-dimensional reconstruction of the small enzyme from negatively stained membrane crystals, also described here, shows only a peripheral domain. These results suggest that the membrane protruding arm of the large form corresponds to the small enzyme, whereas the arm lying within the membrane can be identified as the hydrophobic fragment. The two parts of NADH dehydrogenase that can be defined by the separate genetic origin of (most of) their subunits, their independent assembly, and their distinct contributions to the electron pathway can thus be assigned to the two arms of the L-shaped complex I.