Structure Of Integral Membrane Proteins Research Articles

Amino acid distributions in integral membrane protein structures

Advances in structure determination of membrane proteins enable analysis of the propensities of amino acids in extramembrane versus transmembrane locations to be performed on the basis of structure rather than of sequence and predicted topology. Using 29 available structures of integral membrane proteins with resolutions better than 4 Å the distributions of amino acids in the transmembrane domains were calculated. The results were compared to analysis based on just the sequences of the same transmembrane α-helices and significant differences were found. The distribution of residues between transmembrane α-helices and β-strands was also compared. Large hydrophobic (Phe, Leu, Ile, Val) residues showed a clear preference for the protein surfaces facing the lipids for β-barrels, but in α-helical proteins no such preference was seen, with these residues equally distributed between the interior and the surface of the protein. A notable exception to this was alanine, which showed a slight preference for the interior of α-helical membrane proteins. Aromatic residues were found to follow saddle-like distributions preferring to be located in the lipid/water interfaces. The resultant ‘aromatic belts’ were spaced more closely for β-barrel than for α-helical membrane proteins. Charged residues could be shown to generally avoid surfaces facing the bilayer although they were found to occur frequently in the transmembrane region of β-barrels. Indeed detailed comparison between α-helical and β-barrel proteins showed many qualitative differences in residue distributions. This suggests that there may be subtle differences in the factors stabilising β-barrels in bacterial outer membranes and α-helix bundles in all other membranes.

K+ channels lacking the 'tetramerization' domain: implications for pore structure.

The difficulty of obtaining high-resolution structures of integral membrane proteins has been a frustrating barrier to understanding the membrane-based functions of living cells. The mere handful of such structures stands out in dismal contrast to the cornucopia of water-soluble proteins comprehensible at the atomic level. Nevertheless, crystallographically tractable preparations of aqueous domains of membrane proteins have provided molecular insight into phenomena as varied as chemotaxis, immune cell responses to antigens, viral infectivity and cellular synthesis of ATP. Recently, the first structural glimpse of a neuronal ion channel was reported - the T1, or 'tetramerization,' domain of a Shaker-type voltage-gated K+ channel at 1.6 A resolution. The isolated domain associates into a water-soluble four-fold symmetric homotetramer. This structure prompted the novel, provocative proposal that the T1 domain is an essential component of the ion permeation pathway, forming a previously unsuspected ion-coordinating constriction on the cytoplasmic side of the channel and acting as the receptor for the pore-blocking 'ball and chain' inactivation peptide. It has also been commonly conjectured that the T1 domain is required for tetramerization in the channel maturation process. By studying the detailed properties of Shaker K+ channels in which the T1 domain is deleted, we show all these proposals to be invalid. The structure of the T1 domain expressed in isolation is therefore unlikely to mirror in detail its structure when attached to the ion-conducting channel.

Effect of protein hydration on receptor conformation: decreased levels of bound water promote metarhodopsin II formation.

Neutral solutes were used to investigate the effects of osmotic stress both on the ability of rhodopsin to undergo its activating conformation change and on acyl chain packing in the rod outer segment (ROS) disk membrane. The equilibrium concentration of metarhodopsin II (MII), the conformation of photoactivated rhodopsin, which binds and activates transducin, was increased by glycerol, sucrose, and stachyose in a manner which was linear with osmolality. Analysis of this shift in equilibrium in terms of the dependence of ln(Keq) on osmolality revealed that 20 +/- 1 water molecules are released during the MI-to-MII transition at 20 degrees C, and at 35 degrees C 13 +/- 1 waters are released. At 35 degrees C the average time constant for MII formation was increased from 1.20 +/- 0.09 ms to 1.63 +/- 0.09 ms by addition of 1 osmolal sucrose or glycerol. The effect of the neutral solutes on acyl chain packing in the ROS disk membrane was assessed via measurements of the fluorescence lifetime and anisotropy decay of 1,6-diphenyl-1,3,5-hexatriene (DPH). Analysis of the anisotropy decay of DPH in terms of the rotational diffusion model showed that the angular width of the equilibrium orientational distribution of DPH about the membrane normal was progressively narrowed by increased osmolality. The parameter fv, which is proportional to the overlap between the DPH orientational probability distribution and a random orientational distribution, was reduced by the osmolytes in a manner which was linear with osmolality. This study highlights the potentially opposing interplay between the effect of membrane surface hydration on both the lipid bilayer and integral membrane protein structure. Our results further demonstrate that the binding and release of water molecules play an important role in modulating functional conformational changes for integral membrane proteins, as well as for soluble globular proteins.

Structures of membrane proteins determined at atomic resolution.

Following determination of the first crystal structure of the reaction center of Rhodopseudomonas viridis, a membrane protein, by X-ray crystal structure analysis at 3.0 A resolution, 18 X-ray crystal structures and two electron crystal structures of membrane proteins have been obtained at higher than 3.5 A resolution. Besides these integral membrane protein structures, three crystal structures of water-soluble proteins, which can enter membranes, have been determined by X-ray crystallography at high resolution. The structural features of membrane proteins have been summarized by inspecting these crystal structures. The polypeptide chain crosses the membrane in a helical conformation or a beta-strand. The central +10 A region of the transmembrane alpha-helix is dominated by hydrophobic residues. On both sides of the central region are concentrated polar aromatic residues. Charged residues are dominant around +15 A to +20 A. All the transmembrane beta-structures are found in pore-forming proteins. The central region of the transmembrane beta-structure is amphipathic with hydrophobic residues on the membrane exposed side. The distribution of amino acid residues on the membrane exposed surface of the transmembrane beta-structure is similar to that of the transmembrane alpha-helix. alpha-Helices anchoring the membrane surface region are amphipathic with hydrophobic residues inside and hydrophilic residues outside.

Structural analysis of cloned plasma membrane proteins by freeze-fracture electron microscopy.

We have used freeze-fracture electron microscopy to examine the oligomeric structure and molecular asymmetry of integral plasma membrane proteins. Recombinant plasma membrane proteins were functionally expressed in Xenopus laevis oocytes, and the dimensions of their freeze-fracture particles were analyzed. To characterize the freeze-fracture particles, we compared the particle cross-sectional area of proteins with alpha-helical transmembrane domains (opsin, aquaporin 1, and a connexin) with their area obtained from existing maps calculated from two-dimensional crystals. We show that the cross-sectional area of the freeze-fracture particles corresponds to the area of the transmembrane domain of the protein, and that the protein cross-sectional area varies linearly with the number membrane-spanning helices. On average, each helix occupies 1.40 +/- 0.03 nm2. By using this information, we examined members from three classes of plasma membrane proteins: two ion channels, the cystic fibrosis transmembrane conductance regulator and connexin 50 hemi-channel; a water channel, the major intrinsic protein (the aquaporin 0); and a cotransporter, the Na+/glucose cotransporter. Our results suggest that the cystic fibrosis transmembrane conductance regulator is a dimer containing 25 +/- 2 transmembrane helices, connexin 50 is a hexamer containing 24 +/- 3 helices, the major intrinsic protein is a tetramer containing 24 +/- 3 helices, and the Na+/glucose cotransporter is an asymmetrical monomer containing 15 +/- 2 helices.

Stoichiometry of lipid-protein interaction and integral membrane protein structure

The stoichiometries of lipid-protein interaction obtained from spin label electron spin resonance experiments with integral membrane proteins are compared with simple geometric models for the intramembranous perimeter that are based on the predicted numbers of transmembrane helices. Deviations from the predicted values provide evidence for oligomerization of the protein in the membrane and/or more complex arrangements of the transmembrane segments.

A Model Recognition Approach to the Prediction of All-Helical Membrane Protein Structure and Topology

This paper describes a new method for the prediction of the secondary structure and topology of integral membrane proteins based on the recognition of topological models. The method employs a set of statistical tables (log likelihoods) complied from well-characterized membrane protein data, and a novel dynamic programming algorithm to recognize membrane topology models by expectation maximization. The statistical tables show definite biases toward certain amino acid species on the inside, middle, and outside of a cellular membrane. Using a set of 83 integral membrane protein sequences taken from a variety of bacterial, plant, and animal species, and a strict jackknifing procedure, where each protein (along with any detectable homologues) is removed from the training set used to calculate the tables before prediction, the method successfully predicted 64 of the 83 topologies, and of the 37 complex multispanning topologies 34 were predicted correctly.

Integral membrane protein structure: transmembrane α-helices as autonomous folding domains: Current opinion in structural biology 1993, 3: 532–540

The transmembrane region of many integral membrane proteins is made up of a bundle of hydrophobic α-helices. Such a structure could result from a two-stage folding process, during which preformed transmembrane helices with independent stability pack without topological rearrangement. This view was originally prompted by experiments in which fragments of transmembrane regions were separately refolded into lipid bilayers and subsequently brought together to yield a functional protein. Other lines of evidence, including the existence of ‘one-helix’ miniproteins, gene-fusion experiments, helix-driven oligomerization of bitopic proteins, and sequence rearrangements in the course of evolution support this view. Although it forms a useful basis for structural predictions, the limitations of the two-stage folding hypothesis are not clearly defined, and the proportion of integral membrane proteins to which it applies remains uncertain. The papers discussed in the present review illustrate recent progress along these lines.

Effects of Cycloheximide and Tunicamycin on Lysosomal Cystine Transport in Rat FRTL-5 Cells

Rat thyroid FRTL-5 cells were employed to study the synthesis and degradation of a functional integral lysosomal membrane protein, the lysosomal cystine transporter. This carrier exhibited countertransport and closely resembled the cystine transport system of human leucocytes, fibroblasts, and lymphoblasts shown to be defective in the lysosomal storage disease, nephropathic cystinosis. Using cycloheximide to prevent new protein synthesis, the half-life of the FRTL-5 cell lysosomal cystine carrier was determined to approximate 21 h. Carrier function was not influenced by the N-glycosylation inhibitor tunicamycin, nor by the oligosaccharide processing inhibitors castanospermine and deoxymannojirimycin. The data suggest that the lysosomal cystine carrier is a protein without strict functional requirements for N-linked oligosaccharides, and that rat FRTL-5 cells can be employed in future investigations into the structure and function of other integral lysosomal membrane proteins as well.

Structural Studies of the Enveloped dsRNA Bacteriophage θof Pseudomonas syringae by Raman Spectroscopy: I. The Virion and Its Membrane Envelope

We report and interpret the first Raman spectrum of a double-stranded RNA virus containing a membrane envelope. Spectra of the native bacteriophage θ6 and of its isolated host-attachment (spike) protein and phospholipid-free core assembly were collected from aqueous solutions over a wide range of temperature. Comparison of the vibrational spectra by digital difference methods permits the following structural conclusions regarding molecular constituents of the fully assembled virion. (1) The double-stranded RNA, phospholipid and protein components of the phage exhibit Raman amplitudes in accordance with their biochemically determined compositions in the native virion (10, 20 and 70%, respectively). (2) α-Helix and irregular conformations are the dominant secondary structures in proteins of both the viral membrane and nucleocapsid. This represents a departure from previously examined icosahedral phage and plant viruses, which are dominated by β-sheet structures. (3) The phospholipids of the viral membrane are liquid crystalline throughout the determined range of virus thermostability (0 to 40°C). (4) The P3 spike protein of θ6, which is anchored to, but not sequestered within the viral membrane, is largely α-helical (∼35%) and highly thermolabile. Denaturation of P3 at temperatures above 30°C leads to appreciable loss (∼ 20%) of α-helix in favor of β-strand structure, and alters significantly the environments of many aromatic side-chains. (5) The secondary structures of integral membrane proteins of θ6 are overwhelmingly α-helical (∼70 to 80%) and also thermolabile. In contrast to P3, which exhibits aspartate and glutamate carboxyls in the ionized form (CO-2), the integral membrane proteins exhibit only protonated carboxyl groups (COOH).Treatment of θ6 with butylated hydroxytoluene (BHT), which has been shown to remove the P3 spike protein, does not significantly perturb phospholipids and associated integral proteins of the viral membrane or structural proteins and packaged double-stranded RNA of the nucleocapsid. However, P3 subunits, which are recovered after BHT treatment, exhibit radically altered secondary and tertiary structures, including the loss of most subunit α-helices. Among the P3 side-chains affected by BHT treatment, we note a general trend toward greater hydrophilicity and greater solvent exposure of the aromatic residues Trp and Tyr. On the other hand, the cysteine sulfhydryl groups of the BHT-isolated P3 monomer are not solvent exposed and function as strong hydrogen-bond donors in the protein core. The BHT-mediated structure transition of P3 is discussed in relation to putative quaternary rearrangements involving the membrane anchor protein, P6, and homomultimer complexes of the P3 monomer.The present studies show that RNA, nucleocapsid and viral membrane components of an enveloped double-stranded RNA virus can be distinguished and structurally characterized by Raman spectroscopy. Importantly, we have demonstrated the ability to obtain Raman spectra from the intact viral membrane, establishing the feasibility of Raman spectroscopy for structural analysis of protein molecules embedded in a viral phospholipid bilayer. The results obtained also serve as the basis for detailed spectroscopic analyses of the θ6 nucleocapsid, the θ6 polymerase complex (nucleocapsid core), and the packaged θ6 genome, as described in the accompanying paper (Bamford et al., 1993).

Solid-state NMR approaches for studying membrane protein structure.

The proteins that span the cell membrane perform an array of functions ranging from signal transduction to ion transport. In the past few years, gene sequencing of membrane proteins has yielded an abundance of amino acid sequence data that in turn has generated structural models for many of these proteins (13, 14,52,61). Unfortunately, high resolution structures of integral membrane proteins in phospholipid bilayers have been notori­ ously difficult to obtain using either X-ray diffraction or solution NMR methods. It has generally been a challenge to grow well-ordered crystals of large hydrophobic molecules for diffraction studies, while solution NMR

A murine T lymphocyte antigen belongs to a supergene family of type II integral membrane proteins.

A murine cell surface, disulfide-linked 85kDa dimer, defined with murine mAb A1, is expressed at high levels on EL-4 cells, but at low levels on normal C57BL/6 T cells. A similar structure is recognized by the rat mAbs YE1/32 and YE1/48. We isolated a cDNA clone encoding the antigen recognized by mAb A1 by immunoselection of a cDNA library in the eukaryotic expression vector CDM8. COS 7.2 cells transfected with this cDNA clone expressed an mAb A1-reactive 85 kDa disulfide-linked dimer with 44 kDa subunits, which was also reactive with the mAbs YE1/32 and YE1/48. The A1 gene displayed extensive strain polymorphism, underwent no rearrangement in EL-4, and hybridized with multiple restriction fragments, suggesting that it is a member of a multi-gene family. The deduced polypeptide contained 262 residues with an m.w. of 30,648, multiple cysteines, and three potential N-linked glycosylation sites, consistent with previous observations. In contrast to most integral membrane proteins, the putative A1 protein had features of a type II integral membrane protein structure, with its carboxyl terminus exposed extracellularly and an intracytoplasmic amino terminus. There was significant homology with several type II integral membrane proteins, including the human and chicken asialoglycoprotein receptors, and especially the human low affinity Fc epsilon receptor, in the putative extracellular domains of these proteins. This analysis suggested that the A1 gene belongs to a novel supergene family of type II integral membrane proteins and suggested that the A1 protein itself may be involved in binding a soluble ligand such as carbohydrates or immunoglobulin.

Effect of detergents on the structure of integral membrane proteins of sendai virus studied with size-exclusion high-performance liquid chromatography and monoclonal antibodies

The integral membrane proteins of Sendai virus, the fusion protein F ( M r = 65 000) and the haemagglutinin-neuraminidase protein HN ( M r = 68 000), were used as a model protein mixture. They were subjected to size-exclusion high-performance liquid chromatography on Superose 6HR columns with eluents containing various additives in order to solubilize the proteins. The effect of the additives on the structure of the membrane proteins was investigated with conformation-dependent monoclonal antibodies, either directed against F or HN protein, and by determination of the haemagglutinating capacity of the HN protein. The results show that the structure of the HN protein is more easily disturbed by eluents than that of the F protein. When the elution conditions are mild, e.g., 0.1% octylglucoside, the structure of both proteins is conserved but no separation is obtained. Elution with a buffer containing 0.05% sarkosyl (dodecyl methylglycine sodium salt) did not affect the structure and resulted in pure F protein. Pretreatment of the Amberlite XAD-2-treated Sendai virus envelope extract with 4% sodium dodecyl sulphate (SDS) and elution with 0.1% SDS in 50 m M sodium phosphate (pH 6.5) altered the structure of the HN protein but resulted in purification of the tetramer and the dimer of the HN protein, and the monomer of the F protein.

The growth and characterization of membrane protein crystals

A major advance in the study of integral membrane protein structure has been the development of methods for crystallizing these amphiphilic protein species. The crystals were obtained from isotropic solutions of protein and nonionic detergents and contain a substantial amount of detergent bound within the crystal lattice. Standard techniques for crystallizing water-soluble proteins can be applied to membrane proteins if the physical characteristics and behavior of the detergent system used for protein solubilization are adequately controlled. In this report we present the results of some crystallization experiments on porin, a protein forming transmembrane channels in the outer membrane of E. coli, and discuss the detergent-related phenomena which seem to affect the crystallization process.

The Interaction between a Synthetic Amphiphilic Polypeptide and Lipids

Most spectroscopic studies designed to examine the interactions between phospholipid and protein molecules in biological membranes have involved measurements of the perturbation of the orientational order of lipids due to the presence of proteins (1). Such measurements are difficult to interpret on the molecular level because of the complexity of proteins and the paucity of experimental information on the three-dimensional structure of integral membrane proteins. In addition, the proteins are often relatively short lived under the variety of experimental conditions required for physical measurements and a large biochemical effort is required to produce an adequate renewable supply of well-characterized integral membrane proteins. For these reasons, we have been led to develop a novel class of model membranes which have properties similar to those of protein-lipid reconstituents and which are stable and well-defined physical systems. Our model membrane consists of amphiphilic polypeptide molecules incorporated in phospholipid bilayers. The polypeptide is synthesized with hydrophilic amino acids at each end and a number (N) of hydrophobic amino acids in the central part to span the hydrophobic region of the phospholipid bilayer. From the structure of bacteriorhodopsin and general energy considerations of polypeptides in hydrophobic solvents (2), it is anticipated that the most stable type of protein secondary structural element in membranes should be the a-helix, which has a length of 1.5 A/residue. We report here a study of a mixture of the amphilic polypeptide (Lys2-Leu24-Lys2-Ala amide) with dipalmitoylphosphatidylcholine (DPPC) in excess water (50% by weight). Since the hydrophobic region of DPPC has a width of 33 A in the liquid crystalline (La) phase and 47 A in the gel (Lo phase (reference 1, page 38) we have decided to study these lipid/polypeptide mixtures as a function of values of N in the range 20-30 and of molar ratio (L/P). This paper should be considered as a first report of a systematic study.

Thermodynamics, the structure of integral membrane proteins, and transport.

Membranes are structures whose lipid and protein components are at, or close to, equilibrium in the plane of the membrane, but are not at equilibrium across the membrane. The thermodynamic tendency of ionic and highly polar molecules to be in contact with water rather than with nonpolar media (hydrophilic interactions) is important in determining these equilibrium and nonequilibrium states. In this paper, we speculate about the structures and orientations of integral proteins in a membrane, and about how the equilibrium and nonequilibrium features of such structures and orientations might be influenced by the special mechanisms of biosynthesis, processing, and membrane insertion of these proteins. The relevance of these speculations to the mechanisms of the translocation event in membrane transport is discussed, and specific protein models of transport that have been proposed are analyzed.