Computational searching and mutagenesis suggest a structure for the pentameric transmembrane domain of phospholamban

Computation and mutagenesis suggest a right-handed structure for the synaptobrevin transmembrane dimer

Biological membrane fusion involves a highly precise and ordered set of protein-protein interactions. Synaptobrevin is a key player in this process. Mutagenesis studies of its single transmembrane segment suggest that it dimerizes in a sequence specific manner. Using the computational methods developed for the successful structure prediction of the glycophorin A transmembrane dimer, we have calculated a structural model for the synaptobrevin dimer. Our computational search yields a well-populated cluster of right-handed structures consistent with the experimentally determined dimerization motif. The three-dimensional structure contains an interface formed primarily by leucine and isoleucine side-chain atoms and has no interhelical hydrogen bonds. The model is the first three-dimensional picture of the synaptobrevin transmembrane dimer and provides a basis for further focused experimentation on its structure and association thermodynamics.

Substitution rates in alpha-helical transmembrane proteins

It has been shown previously that some membrane proteins have a conserved core of amino acid residues. This idea not only serves to orient helices during model building exercises but may also provide insight into the structural role of residues mediating helix-helix interactions. Using experimentally determined high-resolution structures of alpha-helical transmembrane proteins we show that, of the residues within the hydrophobic transmembrane spans, the residues at lipid and subunit interfaces are more evolutionarily variable than those within the lipid-inaccessible core of a polypeptide's transmembrane domain. This supports the idea that helix-helix interactions within the same polypeptide chain and those at the interface between different polypeptide chains may arise in distinct ways. To show this, we use a new method to estimate the substitution rate of an amino acid residue given an alignment and phylogenetic tree of closely related proteins. This method gives better sensitivity in the otherwise-conserved transmembrane domains than a conventional similarity analysis and is relatively insensitive to the sequences used.

Mapping the energy surface of transmembrane helix-helix interactions

Transmembrane helices are no longer believed to be just hydrophobic segments that exist solely to anchor proteins to a lipid bilayer, but rather they appear to have the capacity to specify function and structure. Specific interactions take place between hydrophobic segments within the lipid bilayer whereby subtle mutations that normally would be considered innocuous can result in dramatic structural differences. That such specificity takes place within the lipid bilayer implies that it may be possible to identify the most favorable interaction surface of transmembrane alpha-helices based on computational methods alone, as shown in this study. Herein, an attempt is made to map the energy surface of several transmembrane helix-helix interactions for several homo-oligomerizing proteins, where experimental data regarding their structure exist (glycophorin A, phospholamban, Influenza virus A M2, Influenza virus C CM2, and HIV vpu). It is shown that due to symmetry constraints in homo-oligomers the computational problem can be simplified. The results obtained are mostly consistent with known structural data and may additionally provide a view of possible alternate and intermediate configurations.

Role of cysteine residues in structural stability and function of a transmembrane helix bundle

To study the structural and functional roles of the cysteine residues at positions 36, 41, and 46 in the transmembrane domain of phospholamban (PLB), we have used Fmoc (N-(9-fluorenyl)methoxycarbonyl) solid-phase peptide synthesis to prepare alpha-amino-n-butyric acid (Abu)-PLB, the analogue in which all three cysteine residues are replaced by Abu. Whereas previous studies have shown that replacement of the three Cys residues by Ala (producing Ala-PLB) greatly destabilizes the pentameric structure, we hypothesized that replacement of Cys with Abu, which is isosteric to Cys, might preserve the pentameric stability. Therefore, we compared the oligomeric structure (from SDS-polyacrylamide gel electrophoresis) and function (inhibition of the Ca-ATPase in reconstituted membranes) of Abu-PLB with those of synthetic wild-type PLB and Ala-PLB. Molecular modeling provides structural and energetic insight into the different oligomeric stabilities of these molecules. We conclude that 1) the Cys residues of PLB are not necessary for pentamer formation or inhibitory function; 2) the steric properties of cysteine residues in the PLB transmembrane domain contribute substantially to pentameric stability, whereas the polar or chemical properties of the sulfhydryl group play only a minor role; 3) the functional potency of these PLB variants does not correlate with oligomeric stability; and 4) acetylation of the N-terminal methionine has neither a functional nor a structural effect in full-length PLB.

Genetic selection for and molecular dynamic modeling of a protein transmembrane domain multimerization motif from a random Escherichia coli genomic library

In order to identify new transmembrane helix packing motifs in naturally occurring proteins, we have selected transmembrane domains from a library of random Escherichia coli genomic DNA fragments and screened them for homomultimerization via their abilities to dimerize the bacteriophage lambda cI repressor DNA-binding domain. Sequences were isolated using a modified lambda cI headpiece dimerization assay system, which was shown previously to measure transmembrane helix-helix association in the E. coli inner membrane. Screening resulted in the identification of several novel sequences that appear to mediate helix-helix interactions. One sequence, representing the predicted sixth transmembrane domain (TM6) of the E. coli protein YjiO, was chosen for further analysis. Using site-directed mutagenesis and molecular dynamics, a small set of models for YjiO TM6 multimerization interface interactions were generated. This work demonstrates the utility of combining in vivo genetic tools with computational systems for understanding membrane protein structure and assembly.

Sarcolipin, the shorter homologue of phospholamban, forms oligomeric structures in detergent micelles and in liposomes

The human 31-amino acid integral membrane protein sarcolipin (SLN), which regulates the sarcoplasmic reticulum Ca-ATPase in fast-twitch skeletal muscle, was chemically synthesized. Appropriate synthesis and purification strategies were used to achieve high purity and satisfactory yields of this hydrophobic and poorly soluble protein. Structural and functional properties of SLN were analyzed and compared with the homologous region of human phospholamban (PLB) comprising residues Ala(24)-Leu(52) (PLB-(24-52)), the regulatory protein of the cardiac sarcoplasmic reticulum Ca-ATPase. Circular dichroism spectroscopy showed that SLN is a predominantly alpha-helical protein and that the secondary structure is highly resistant to SDS and thermal denaturation. In this respect SLN is remarkably similar to PLB-(24-52). However, SLN is monomeric in SDS gels, whereas PLB-(24-52) shows a monomer-pentamer equilibrium typical for native PLB. Analytical ultracentrifugation experiments revealed that SLN oligomerizes in the presence of the nonionic detergents octylpolyoxyethylene and octyl glucoside in a concentration-dependent manner. No plateau was observed, and a pentameric state was only reached at much higher protein concentrations compared with PLB-(24-52). Chemical cross-linking showed that also in liposomes SLN has the ability to self-associate to oligomers. PLB-(24-52) specifically oligomerized to pentamers in the presence of octylpolyoxyethylene as well as in liposomes at low protein concentrations. In the presence of octylpolyoxyethylene pentamers were the main oligomeric species, whereas in liposomes monomers and dimers were predominant. Increasing the protein concentration led to self-association of PLB-(24-52) pentamers in the presence of octylpolyoxyethylene. Functional reconstitution of Ca-ATPase with PLB-(24-52) and SLN in liposomes showed that both proteins regulate the Ca-ATPase in a similar manner.

A new method to model membrane protein structure based on silent amino acid substitutions

The importance of accurately modeling membrane proteins cannot be overstated, in lieu of the difficulties in solving their structures experimentally. Often, however, modeling procedures (e.g., global searching molecular dynamics) generate several possible candidates rather then pointing to a single model. Herein we present a new approach to select among candidate models based on the general hypothesis that silent amino acid substitutions, present in variants identified from evolutionary conservation data or mutagenesis analysis, do not affect the stability of a native structure but may destabilize the non-native structures also found. The proof of this hypothesis has been tested on the alpha-helical transmembrane domains of two homodimers, human glycophorin A and human CD3-zeta, a component of the T-cell receptor. For both proteins, only one structure was identified using all the variants. For glycophorin A, this structure is virtually identical to the structure determined experimentally by NMR. We present a model for the transmembrane domain of CD3-zeta that is consistent with predictions based on mutagenesis, homology modeling, and the presence of a disulfide bond. Our experiments suggest that this method allows the prediction of transmembrane domain structure based only on widely available evolutionary conservation data.

The structure of the HIV-1 Vpu ion channel: modelling and simulation studies

Vpu is an 81 amino acid auxiliary protein in HIV-1 which exhibits channel activity. We used two homo-pentameric bundles with the helical transmembrane segments derived from FTIR spectroscopy in combination with a global molecular dynamics search protocol: (i) tryptophans (W) pointing into the pore, and (ii) W facing the lipids. Two equivalent bundles have been generated using a simulated annealing via a restrained molecular dynamics simulations (SA/MD) protocol. A fifth model was generated via SA/MD with all serines facing the pore. The latter model adopts a very stable structure during the 2 ns of simulation. The stability of the models with W facing the pore depends on the starting structure. A possible gating mechanism is outlined.

Helical membrane protein folding, stability, and evolution

Helical membrane protein folding and oligomerization can be usefully conceptualized as involving two energetically distinct stages-the formation and subsequent side-to-side association of independently stable transbilayer helices. The interactions of helices with the bilayer, with prosthetic groups, and with each other are examined in the context of recent evidence. We conclude that the two-stage concept remains useful as an approach to simplifying discussions of stability, as a framework for folding concepts, and as a basis for understanding membrane protein evolution.

Coiled coils: a highly versatile protein folding motif

The alpha-helical coiled coil is one of the principal subunit oligomerization motifs in proteins. Its most characteristic feature is a heptad repeat pattern of primarily apolar residues that constitute the oligomer interface. Despite its simplicity, it is a highly versatile folding motif: coiled-coil-containing proteins exhibit a broad range of different functions related to the specific 'design' of their coiled-coil domains. The architecture of a particular coiled-coil domain determines its oligomerization state, rigidity and ability to function as a molecular recognition system. Much progress has been made towards understanding the factors that determine coiled-coil formation and stability. Here we discuss this highly versatile protein folding and oligomerization motif with regard to its structural architecture and how this is related to its biological functions.

The Croonian Lecture 2000. Nicotinic acetylcholine receptor and the structural basis of fast synaptic transmission

Communication in the nervous system takes place at chemical and electrical synapses, where neurotransmitter-gated ion channels, such as the nicotinic acetylcholine (ACh) receptor, and gap junction channels control propagation of electrical signals from one cell to the next. Newly developed electron crystallographic methods have revealed the structures of these channels trapped in open as well as closed states, suggesting how they work. The ACh receptor has large vestibules extending from the membrane which shape the ACh-binding pockets and facilitate selective transport of cations across a narrow membrane-spanning pore. When ACh enters the pockets it triggers a concerted conformational change that opens the pore by destabilizing a gate in the middle of the membrane made by a ring of pore-lining alpha-helical segmets. The alternative 'open' configuration of pore-lining segments reshapes the lumen and creates new surfaces, allowing the ions to pass through. The gap junction channel uses a similar structural mechanism, involving coordinated rearrangements of alpha-helical segments in the plane of the membrane, to open its pore.

Use of a new label, (13)==(18)O, in the determination of a structural model of phospholamban in a lipid bilayer. Spatial restraints resolve the ambiguity arising from interpretations of mutagenesis data

A structural model of pentameric phospholamban (Plb) in a lipid bilayer has been derived using a combination of experimental data, obtained from ATR-FTIR site-directed dichroism, and the implementation of the resulting restraints during a molecular dynamics simulation. Plb (residues 24-52) has been synthesised incorporating a new label, 1-(13)C==(18)O, at residues 42 and 43. We have not only determined the tilt of the helices, 10(+/-6) degrees, but also the relative orientation of the transmembrane segments, with an omega angle of -32(+/-10) degrees for L42. This angle is taken as zero in the direction of the helix tilt. Plb is a simple test case where site-directed dichroism has been applied to resolve the indeterminacy arising from the mutagenesis data available. The results presented point specifically to a single structural model for Plb.

Recursive use of evolutionary conservation data in molecular modeling of membrane proteins A model of the multidrug H+ antiporter emrE

Membrane proteins are currently the most biomedically important family of proteins, serving as targets for the majority of pharmaceutical agents. It is also clear that they are invariably abundant in all of the genomes sequence so far, representing up to a third of all open reading frames. Finally, and regrettably, it is clear that they are highly resistant to structural elucidation, representing less than 0.2% of the Protein Data Bank. Recent accomplishments in genome sequencing efforts, however, may help offset this imbalance through the availability of evolutionary conservation data. Herein, we develop a novel approach, utilizing a combination of evolutionary conservation data and global searching molecular dynamics simulations to model membrane proteins, deriving a model for the multidrug H+ antiporter EmrE, a transmembrane four-helix bundle. Structures resulting from an extensive, rotational molecular dynamics search, were evaluated by comparing the residue specific interaction energy and the evolutionary conservation data. Subsequent rounds of molecular dynamics, in which confinement of the search space was undertaken in order to achieve a self consistent result, point to a structure that best satisfies the evolutionary conservation data. As the conservation patterns calculated for each of the helices suggested that the different conservation pattern for helix 3 (as well as being the most conserved) might be due to the oligomeric nature of EmrE, a dodecamer of helices was constructed based on the result of a search of helix 3 as a trimer. The resulting interaction energy per residue in the final model is in reasonable agreement with the evolutionary data and consistent with recent site directed mutagenesis experiments, pointing to the strength of this method as a general tool.

Peptide mimics of the M13 coat protein transmembrane segment. Retention of helix-helix interaction motifs

Sequence-specific noncovalent helix-helix interactions between transmembrane (TM) segments in proteins are investigated by incorporating selected TM sequences into synthetic peptides using the construct CKKK-TM-KKK. The peptides are of suitable hydrophobicity for spontaneous membrane insertion, whereas formation of an N-terminal S-S bond can bring pairs of TM helices into proximity and promote their parallel orientation. Using the propensity of the protein to undergo thermally induced alpha-helix --> beta-sheet transitions as a parameter for helix stability, we compared the wild type and mutant (V29A and V31A) bacteriophage M13 coat proteins with their corresponding TM peptide constructs (M13 residues 24-42). Our results demonstrated that the relevant helix-helix tertiary contacts found in the intact proteins persist in the peptide mimics. Molecular dynamics simulations support the tight "two in-two out" dimerization motif for V31A consistent with mutagenesis data. The overall results reinforce the notion of TM segments as autonomous folding domains and suggest that the generic peptide construct provides a viable reductionist system for membrane protein structural and computational analysis.

Structure of the influenza C virus CM2 protein transmembrane domain obtained by site-specific infrared dichroism and global molecular dynamics searching

The 115-residue protein CM2 from Influenza C virus has been recently characterized as a tetrameric integral membrane glycoprotein. Infrared spectroscopy and site-directed infrared dichroism were utilized here to determine its transmembrane structure. The transmembrane domain of CM2 is alpha-helical, and the helices are tilted by beta = (14.6 +/- 3.0) degrees from the membrane normal. The rotational pitch angle about the helix axis omega for the 1-(13)C-labeled residues Gly(59) and Leu(66) is omega = (218 +/- 17) degrees, where omega is defined as zero for a residue pointing in the direction of the helix tilt. A detailed structure was obtained from a global molecular dynamics search utilizing the orientational data as an energy refinement term. The structure consists of a left-handed coiled-coil with a helix crossing angle of Omega = 16 degrees. The putative transmembrane pore is occluded by the residue Met(65). In addition hydrogen/deuterium exchange experiments show that the core is not accessible to water.

Helix-bundle membrane protein fold templates

In the fold recognition approach to structure prediction, a sequence is tested for compatibility with an already known fold. For membrane proteins, however, few folds have been determined experimentally. Here the feasibility of computing the vast majority of likely membrane protein folds is tested. The results indicate that conformation space can be effectively sampled for small numbers of helices. The vast majority of potential monomeric membrane protein structures can be represented by about 30-folds for three helices, but increases exponentially to about 1,500,000 folds for seven helices. The generated folds could serve as templates for fold recognition or as starting points for conformational searches that are well distributed throughout conformation space.

Transmembrane helix M6 in sarco(endo)plasmic reticulum Ca(2+)-ATPase forms a functional interaction site with phospholamban. Evidence for physical interactions at other sites

In an earlier study (Kimura, Y., Kurzydlowski, K., Tada, M., and MacLennan, D. H. (1997) J. Biol. Chem. 272, 15061-15064), mutation of amino acids on one face of the phospholamban (PLN) transmembrane helix led to loss of PLN inhibition of sarco(endo)plasmic reticulum Ca(2+)-ATPase (SERCA) molecules. This helical face was proposed to form a site of PLN interaction with a transmembrane helix in SERCA molecules. To determine whether predicted transmembrane helices M4, M5, M6, or M8 in SERCA1a interact with PLN, SERCA1a mutants were co-expressed with wild-type PLN and effects on Ca(2+) dependence of Ca(2+) transport were measured. Wild-type inhibitory interactions shifted apparent Ca(2+) affinity of SERCA1a by an average of -0.34 pCa units, but four of the seven mutations in M4 led to a more inhibitory shift in apparent Ca(2+) affinity, averaging -0.53 pCa units. Seven mutations in M5 led to an average shift of -0.32 pCa units and seven mutations in M8 led to an average shift of -0.30 pCa units. Among 11 mutations in M6, 1, Q791A, increased the inhibitory shift (-0.59 pCa units) and 5, V795A (-0.11), L802A (-0.07), L802V (-0.04), T805A (-0.11), and F809A (-0.12), reduced the inhibitory shift, consistent with the view that Val(795), Leu(802), Thr(805), and Phe(809), located on one face of a predicted M6 helix, form a site in SERCA1a for interaction with PLN. Those mutations in M4, M6, or M8 of SERCA1a that enhanced PLN inhibitory function did not enhance PLN physical association with SERCA1a, but mutants V795A and L802A in M6, which decreased PLN inhibitory function, decreased physical association, as measured by co-immunoprecipitation. In related studies, those PLN mutants that gained inhibitory function also increased levels of co-immunoprecipitation of wild-type SERCA1a and those that lost inhibitory function also reduced association, correlating functional interaction sites with physical interaction sites. Thus, both functional and physical data confirm that PLN interacts with M6 SERCA1a.

A Day in the Life of Dr K. or How I Learned to Stop Worrying and Love Lysozyme: a tragedy in six acts

About the play: In modern drama, the agonizing nature of membrane protein work has not been adequately acknowledged. It is perhaps significant that the first attempt to bring this darker aspect of human existence into focus comes from a Scandinavian author, writing in the tradition of Ibsen and Strindberg but with a distinctly turn-of-the-millenium approach to the inner life of his characters: the despairing Dr K; the cynical Dr R with his post-modernistic life credo; the ambitious but unfeeling Dr C; the modern Ubermensch, Dr B. , with his almost Nietzschean view of human nature. This is a play that is brutally honest, yet full of empathy for the poor souls that get caught between the Scylla of unreachable scientific glory and the Charybdis of helpless mediocrity.James Glib-Burdock, drama critic for The Stratford Observer.

vpu transmembrane peptide structure obtained by site-specific fourier transform infrared dichroism and global molecular dynamics searching

The recently developed method of site-directed Fourier transform infrared dichroism for obtaining orientational constraints of oriented polymers is applied here to the transmembrane domain of the vpu protein from the human immunodeficiency virus type 1 (HIV-1). The infrared spectra of the 31-residue-long vpu peptide reconstituted in lipid vesicles reveal a predominantly alpha-helical structure. The infrared dichroism data of the (13)C-labeled peptide yielded a helix tilt beta = (6.5 +/- 1.7) degrees from the membrane normal. The rotational pitch angle omega, defined as zero for a residue located in the direction of the helix tilt, is omega = (283 +/- 11) degrees for the (13)C labels Val(13)/Val(20) and omega = (23 +/- 11) degrees for the (13)C labels Ala(14)/Val(21). A global molecular dynamics search protocol restraining the helix tilt to the experimental value was performed for oligomers of four, five, and six subunits. From 288 structures for each oligomer, a left-handed pentameric coiled coil was obtained, which best fits the experimental data. The structure reveals a pore occluded by Trp residues at the intracellular end of the transmembrane domain.

Are membrane proteins "inside-out" proteins?

One of the central paradigms of structural biology is that membrane proteins are "inside-out" proteins, in that they have a core of polar residues surrounded by apolar residues. This is the reverse of the characteristics found in water-soluble proteins. We have decided to test this paradigm, now that sufficient numbers of transmembrane alpha-helical structures are accessible to statistical analysis. We have analyzed the correlation between accessibility and hydrophobicity of both individual residues and complete helices. Our analyses reveal that hydrophobicity of residues in a transmembrane helical bundle does not correlate with any preferred location and that the hydrophilic vector of a helix is a poor indicator of the solvent exposed face of a helix. Neither polar nor hydrophobic residues show any bias for the exterior or the interior of a transmembrane domain. As a control, analysis of water-soluble helical bundles performed in a similar manner has yielded clear correlations between hydrophobicity and accessibility. We therefore conclude that, based on the data set used, membrane proteins as "inside-out" proteins is an unfounded notion, suggesting that packing of alpha-helices in membranes is better understood by maximization of van der Waal's forces, rather than by a general segregation of hydrophobicities driven by lipid exclusion.

Activation of the erythropoietin receptor by the gp55-P viral envelope protein is determined by a single amino acid in its transmembrane domain

The spleen focus forming virus (SFFV) gp55-P envelope glycoprotein specifically binds to and activates murine erythropoietin receptors (EpoRs) coexpressed in the same cell, triggering proliferation of erythroid progenitors and inducing erythroleukemia. Here we demonstrate specific interactions between the single transmembrane domains of the two proteins that are essential for receptor activation. The human EpoR is not activated by gp55-P but by mutation of a single amino acid, L238, in its transmembrane sequence to its murine counterpart serine, resulting in its ability to be activated. The converse mutation in the murine EpoR (S238L) abolishes activation by gp55-P. Computational searches of interactions between the membrane-spanning segments of murine EpoR and gp55-P provide a possible explanation: the face of the EpoR transmembrane domain containing S238 is predicted to interact specifically with gp55-P but not gp55-A, a variant which is much less effective in activating the murine EpoR. Mutational studies on gp55-P M390, which is predicted to interact with S238, provide additional support for this model. Mutation of M390 to isoleucine, the corresponding residue in gp55-A, abolishes activation, but the gp55-P M390L mutation is fully functional. gp55-P is thought to activate signaling by the EpoR by inducing receptor oligomerization through interactions involving specific transmembrane residues.

Structure of the 1-36 amino-terminal fragment of human phospholamban by nuclear magnetic resonance and modeling of the phospholamban pentamer

The structure of a 36-amino-acid-long amino-terminal fragment of phospholamban (phospholamban[1-36]) in aqueous solution containing 30% trifluoroethanol was determined by nuclear magnetic resonance. The peptide, which comprises the cytoplasmic domain and six residues of the transmembrane domain of phospholamban, assumes a conformation characterized by two alpha-helices connected by a turn. The residues of the turn are Ile18, Glu19, Met20, and Pro21, which are adjacent to the two phosphorylation sites Ser16 and Thr17. The proline is in a trans conformation. The helix comprising amino acids 22-36 is well determined (the root mean square deviation for the backbone atoms, calculated for a family of 18 nuclear magnetic resonance structures is 0.57 A). Recently, two molecular models of the transmembrane domain of phospholamban were proposed in which a symmetric homopentamer is composed of a left-handed coiled coil of alpha-helices. The two models differ by the relative orientation of the helices. The model proposed by,Simmerman et al. (H.K. Simmerman, Y.M. Kobayashi, J.M. Autry, and L.R. Jones, 1996, J. Biol. Chem. 271:5941-5946), in which the coiled coil is stabilized by a leucine-isoleucine zipper, is similar to the transmembrane pentamer structure of the cartilage oligomeric membrane protein determined recently by x-ray (V. Malashkevich, R. Kammerer, V Efimov, T. Schulthess, and J. Engel, 1996, Science 274:761-765). In the model proposed by Adams et al. (P.D. Adams, I.T. Arkin, D.M. Engelman, and A.T. Brunger, 1995, Nature Struct. Biol. 2:154-162), the helices in the coiled coil have a different relative orientation, i.e., are rotated clockwise by approximately 50 degrees. It was possible to overlap and connect the structure of phospholamban[1-36] derived in the present study to the two transmembrane pentamer models proposed. In this way two models of the whole phospholamban in its pentameric form were generated. When our structure was connected to the leucine-isoleucine zipper model, the inner side of the cytoplasmic domain of the pentamer (where the helices face one another) was lined by polar residues (Gln23, Gln26, and Asn30), whereas the five Arg25 side chains were on the outer side. On the contrary, when our structure was connected to the other transmembrane model, in the inner side of the cytoplasmic domain of the pentamer, the five Arg25 residues formed a highly charged cluster.

Experimentally based orientational refinement of membrane protein models: A structure for the Influenza A M2 H+ channel

The 97-residue M2 protein from Influenza A virus forms H+-selective ion channels which can be attributed solely to the homo-tetrameric alpha-helical transmembrane domain. Site-directed infrared dichroism spectra were obtained for the transmembrane domain of M2, reconstituted in lipid vesicles. Data analysis yielded the helix tilt angle beta=31.6(+/-6.2) degrees and the rotational pitch angle about the helix axis for residue Ala29 omegaAla29=-59.8(+/-9.9) degrees, whereby omega is defined as zero for a residue located in the direction of the helix tilt. A structure was obtained from an exhaustive molecular dynamics global search protocol in which the orientational data are utilised directly as an unbiased refinement energy term. Orientational refinement not only allowed selection of a unique structure but could also be shown to increase the convergence towards that structure during the molecular dynamics procedure. Encouragingly, the structure obtained is highly consistent with all available mutagenesis and conductivity data and offers a direct chemical insight that relates the altered functionality of the channel to its structure.

Two models of the influenza A M2 channel domain: verification by comparison

BACKGROUND

The influenza M2 protein is a simple membrane protein, containing a single transmembrane helix. It is representative of a very large family of single-transmembrane helix proteins. The functional protein is a tetramer, with the four transmembrane helices forming a proton-permeable channel across the bilayer. Two independently derived models of the M2 channel domain are compared, in order to assess the success of applying molecular modelling approaches to simple membrane proteins.

RESULTS

The Calpha RSMD between the two models is 1.7 A. Both models are composed of a left-handed bundle of helices, with the helices tilted roughly 15 degrees relative to the (presumed) bilayer normal. The two models have similar pore radius profiles, with a pore cavity lined by the Ser31 and Gly34 residues and a pore constriction formed by the ring of His37 residues.

CONCLUSIONS

Independent studies of M2 have converged on the same structural model for the channel domain. This model is in agreement with solid state NMR data. In particular, both model and NMR data indicate that the M2 helices are tilted relative to the bilayer normal and form a left-handed bundle. Such convergence suggests that, at least for simple membrane proteins, restraints-directed modelling might yield plausible models worthy of further computational and experimental investigation.

Pentameric assembly of phospholamban facilitates inhibition of cardiac function in vivo

Phospholamban has been proposed to coexist as pentamers and monomers in native sarcoplasmic reticulum membranes. To determine its functional unit in vivo, we reintroduced wild-type (pentameric) or monomeric mutant (C41F) phospholamban in the hearts of phospholamban knockout mice. Transgenic lines, expressing similar levels of mutant or wild-type phospholamban, were identified, and their cardiac phenotypes were characterized in parallel. Sarcoplasmic reticulum Ca2+ transport assays indicated similar decreases in SERCA2 Ca2+ affinity by mutant or wild-type phospholamban. However, the time constants of relaxation and Ca2+ transient decline in isolated cardiomyocytes were diminished to a greater extent by wild-type than mutant phospholamban, even without significant differences in the amplitudes of myocyte contraction and Ca2+ transients between the two groups. Langendorff perfusion also indicated that mutant phospholamban was not capable of depressing the enhanced relaxation parameters of the phospholamban knockout hearts to the same extent as wild-type phospholamban. Moreover, in vivo assessment of mouse hemodynamics revealed a greater depression of cardiac function in wild-type than mutant phospholamban hearts. Thus, the mutant or monomeric form of phospholamban was not as effective in slowing Ca2+ decline or relaxation in cardiomyocytes, hearts, or intact animals as wild-type or pentameric phospholamban. These findings suggest that pentameric assembly of phospholamban is necessary for optimal regulation of myocardial contractility in vivo.

Direct spectroscopic detection of molecular dynamics and interactions of the calcium pump and phospholamban

In order to test molecular models of cardiac calcium transport regulation, we have used spectroscopy to probe the structures, dynamics, and interactions of the Ca pump (Ca-ATPase) and phospholamban (PLB) in cardiac sarcoplasmic reticulum (SR) and in reconstituted membranes. Electron paramagnetic resonance (EPR) and phosphorescence of probes bound to the Ca pump show that the activity of the pump is quite sensitive to its oligomeric interactions. In cardiac SR, PLB aggregates and inhibits the pump, and both effects are reversed by PLB phosphorylation. Previous analyses of PLB's oligomeric state were only in detergent solutions, so we used EPR and fluorescence to determine the oligomeric structure of PLB in its native state in lipid bilayers. Wild-type PLB is primarily oligomeric in the membrane, while the mutant L37A-PLB is monomeric. For both proteins, phosphorylation shifts the dynamic monomer-oligomer equilibrium toward oligomers, and induces a similar structural change, as indicated by tyrosine fluorescence; yet L37A-PLB is more effective than wild-type PLB in inhibiting and aggregating the pump. Fluorescence energy transfer shows that the Ca pump increases the fraction of monomeric PLB, indicating that the pump preferentially binds monomeric PLB. These results support a reciprocal aggregation model for Ca pump regulation, in which the Ca pump is aggregated and inhibited by association with PLB monomers, and phosphorylation of PLB reverses these effects while decreasing the concentration of PLB monomers. To investigate the structure of the PLB pentamer in more detail, we measured the reactivities of cysteine residues in the transmembrane domain of PLB, and recorded EPR spectra of spin labels attached to these sites. These results support an atomic structural model, based on molecular dynamics simulations and mutagenesis studies, in which the PLB pentamer is stabilized by a leucine-isoleucine zipper within the transmembrane domain.

Models for the transmembrane region of the phospholamban pentamer: which is correct?

Phospholamban is a 52-amino acid protein that assembles into a pentamer in the membranes of the sarcoplasmic reticulum. Pentamer formation is driven in large part by interactions of the transmembrane regions of the protein, which are thought to be arranged as interacting alpha-helices. The structural properties of phospholamban have been studied by mutagenesis and optical spectroscopy, resulting in a large database. In this discussion, we present advances in computational modeling, which identifies two probable structures for the transmembrane pentamer. A new approach to mutagenesis is described, which should lead to a clear distinction between the two possible models.

Modelling and simulation of ion channels: applications to the nicotinic acetylcholine receptor

Molecular dynamics simulations with experimentally derived restraints have been used to develop atomic models of M2 helix bundles forming the pore-lining domains of the nicotinic acetylcholine receptor and related ligand-gated ion channels. M2 helix bundles have been used in microscopic simulations of the dynamics and energetics of water and ions within an ion channel. Translational and rotational motion of water are restricted within the pore, and water dipoles are aligned relative to the pore axis by the surrounding helix dipoles. Potential energy profiles for translation of a Na+ ion along the pore suggest that the protein and water components of the interaction energy exert an opposing effect on the ion, resulting in a relatively flat profile which favors cation permeation. Empirical conductance calculations based on a pore radius profile suggest that the M2 helix model is consistent with a single channel conductance of ca. 50 pS. Continuum electrostatics calculations indicate that a ring of glutamate residues at the cytoplasmic mouth of the alpha 7 nicotinic receptor M2 helix bundle may not be fully ionized. A simplified model of the remainder of the channel protein when added to the M2 helix bundle plays a significant role in enhancing the ion selectivity of the channel.

Cardiac protein phosphorylation: functional and pathophysiological correlates

Protein phosphorylation acts a pivotal mechanism in regulating the contractile state of the heart by modulating particular levels of autonomic control on cardiac force/length relationships. Early studies of changes in cardiac protein phosphorylation focused on key components of the excitation-coupling process, namely phospholamban of the sarcoplasmic reticulum and myofibrillar troponin I. In more recent years the emphasis has shifted towards the identification of other phosphoproteins, and more importantly, the delineation of the mechanistic and signaling pathways regulating the various known phosphoproteins. In addition to cAMP- and Ca(2+)-calmodulin-dependent kinase processes, these have included regulation by protein kinase C and the ever-emerging family of growth factor-related kinases such as the tyrosine-, mitogen- and stress-activated protein kinases. Similarly, the role of protein dephosphorylation by protein phosphatases has been recognized as integral in modulating normal cardiac cellular function. Recent studies involving a variety of cardiovascular pathologies have demonstrated that changes in the phosphorylation states of key cardiac regulatory proteins may underlie cardiac dysfunction in disease states. The emphasis of this comprehensive review will be on discussing the role of cardiac phosphoproteins in regulating myocardial function and pathophysiology based not only on in vitro data, but more importantly, from ex vivo experiments with corroborative physiological and biochemical evidence.

Using experimental information to produce a model of the transmembrane domain of the ion channel phospholamban

Molecular models of the transmembrane domain of the phospholamban pentamer have been generated by a computational method that uses the experimentally measured effects of systematic single-site mutations as a guiding force in the modeling procedure. This method makes the assumptions that 1) the phospholamban transmembrane domain is a parallel five-helix bundle, and 2) nondisruptive mutation positions are lipid exposed, whereas 3) disruptive or partially disruptive mutations are not. Our procedure requires substantially less computer time than systematic search methods, allowing rapid assessment of the effects of different experimental results on the helix arrangement. The effectiveness of the approach is investigated in test calculations on two helix-dimer systems of known structure. Two independently derived sets of mutagenesis data were used to define the restraints for generating models of phospholamban. Both resulting models are left-handed, highly symmetrical pentamers. Although the overall bundle geometry is very similar in the two models, the orientation of individual helices differs by approximately 50 degrees, resulting in different sets of residues facing the pore. This demonstrates how differences in restraints can have an effect on the model structures generated, and how the violation of these restraints can identify inconsistent experimental data.

Molecular dynamics simulations of individual alpha-helices of bacteriorhodopsin in dimyristoylphosphatidylcholine. II. Interaction energy analysis

The concepts of hydrophobicity and hydrophobic moments have been applied in attempts to predict membrane protein secondary and tertiary structure. The current paper uses molecular dynamics computer calculations of individual bacteriorhodopsin helices in explicit dimyristoylphosphatidylcholine bilayers to examine the atomic basis of these approaches. The results suggest that the types of interactions between a particular amino acid and the surrounding bilayer depend on the position and type of the amino acid. In particular, aromatic residues are seen to interact favorably at the interface region. Analysis of the trajectories in terms of hydrophobic moments suggests the presence of a particular face that prefers lipid. The results of these simulations may be used to improve secondary structure prediction methods and to provide further insights into the two-stage model of protein folding.

Ion channel stability and hydrogen bonding. Molecular modelling of channels formed by synthetic alamethicin analogues

Several analogues of the channel-forming peptaibol alamethicin have been demonstrated to exhibit faster switching between channel substates than does unmodified alamethicin. Molecular modelling studies are used to explore the possible molecular basis of these differences. Models of channels formed by alamethicin analogues were generated by restrained molecular dynamics in vacuo and refined by short molecular dynamics simulations with water molecules within and at either mouth of the channel. A decrease in backbone solvation was found to correlate with a decrease in open channel stability between alamethicin and an analogue in which all alpha-amino-isobutyric acid residues of alamethicin were replaced by leucine. A decrease in the extent of hydrogen-bonding at residue 7 correlates with lower open channel stabilities of analogues in which the glutamine at position 7 was replaced by smaller polar sidechains. These two observations indicate the importance of alamethicin/water H-bonds in stabilizing the open channel.

Channels formed by the transmembrane helix of phospholamban: a simulation study

Phospholamban is a small membrane protein which can form cation selective ion channels in lipid bilayers. Each subunit contains a single, largely hydrophobic transmembrane helix. The helices are thought to assemble as a pentameric and approximately parallel bundle surrounding a central pore. A model of this assembly (PDB code IPSL) has been used as the starting point for molecular dynamics (MD) simulations of a system consisting of the pentameric helix bundle, plus 217 water molecules located within and at either mouth of the pore. Interhelix distance restraints were employed to maintain the integrity of the helix bundle during a 500 ps MD simulation. Water molecules within the pore exhibited reduced diffusional and rotational mobility. Interactions between the alpha-helix dipoles and the water dipoles, the latter aligned anti-parallel to the former, contribute to the stability of the system. Analysis of the potential energy of interaction of a K+ ion as it was moved through the pore suggested that unfavourable interactions of the cation with the aligned helix dipoles at the N-terminal mouth were overcome by favourable ion-water interactions. Comparable analysis for a Cl ion revealed that the ion-(pore + water) interactions were unfavourable along the whole of the pore, increasingly so from the N- to the C-terminal mouth. Overall, the interaction energy profiles were consistent with a pore selective for cations over anions. Pore radius profiles were used to predict a channel conductance of 50 to 70 ps in 0.2 M KCl, which compares well with an experimental value of 100 ps.

The dielectric properties of water within model transbilayer pores

Ion channels contain extended columns of water molecules within their transbilayer pores. The dynamic properties of such intrapore water have been shown to differ from those of water in its bulk state. In previous molecular dynamics simulations of two classes of model pore (parallel bundles of Ala20 alpha-helices and antiparallel barrels of Ala10 beta-strands), a substantially reduced translational and rotational mobility of waters was observed within the pore relative to bulk water. Molecular dynamics simulations in the presence of a transpore electrostatic field (i.e., a voltage drop along the pore axis) have been used to estimate the resultant polarization (due to reorientation) of the intrapore water, and hence to determine the local dielectric behavior within the pore. It is shown that the local dielectric constant of water within a pore is reduced for models formed by parallel alpha-helix bundles, but not by those formed by beta-barrels. This result is discussed in the context of electrostatics calculations of ion permeation through channels, and the effect of the local dielectric of water within a helix bundle pore is illustrated with a simple Poisson-Boltzmann calculation.

Molecular dynamics of individual alpha-helices of bacteriorhodopsin in dimyristol phosphatidylocholine. I. Structure and dynamics

Understanding the role of the lipid bilayer in membrane protein structure and dynamics is needed for tertiary structure determination methods. However, the molecular details are not well understood. Molecular dynamics computer calculations can provide insight into these molecular details of protein:lipid interactions. This paper reports on 10 simulations of individual alpha-helices in explicit lipid bilayers. The 10 helices were selected from the bacteriorhodopsin structure as representative alpha-helical membrane folding components. The bilayer is constructed of dimyristoyl phosphatidylcholine molecules. The only major difference between simulations is the primary sequence of the alpha-helix. The results show dramatic differences in motional behavior between alpha-helices. For example, helix A has much smaller root-mean-squared deviations than does helix D. This can be understood in terms of the presence of aromatic residues at the interface for helix A that are not present in helix D. Additional motions are possible for the helices that contain proline side chains relative to other amino acids. The results thus provide insight into the types of motion and the average structures possible for helices within the bilayer setting and demonstrate the strength of molecular simulations in providing molecular details that are not directly visualized in experiments.

Helix packing in membrane proteins

A survey of 45 transmembrane (TM) helices and 88 helix packing interactions in three independent transmembrane protein structures reveals the following features. (1) Helix lengths range from 14 to 36 residues with an average length of 26.4 residues. There is a preference for lengths greater than 20 residues. (2) The helices are tilted with respect to the bilayer normal by an average of 21 degrees, but there is a decided preference for smaller tilt angles. (3) The distribution of helix packing angles is very different than for soluble proteins. The most common packing angles for TM helices are centered around +20 degrees while for soluble proteins packing angles of around -35 degrees are the most prevalent. (4) The average distance of closest approach is 9.6 A, which is the same as soluble proteins. (5) There is no preference for the positioning of the point of closest approach along the length of the helices. (6) It is almost a rule that TM helices pack against neighbors in the sequence. Of the 37 helices that have a sequence neighbor, 36 of them are in significant contact with a neighbor. (7) An antiparallel orientation is more prevalent than a parallel orientation and antiparallel interactions are more intimate on average. The general features of helix bundle membrane protein architecture described in this survey should prove useful in the modeling of helix bundle transmembrane proteins.

Helix-helix packing in a membrane-like environment

The unique ability of the glycophorin A transmembrane helix to dimerize in SDS has previously been exploited in studies of the sequence specificity of helix-helix packing in a micellar environment. Here, we have made different insertion mutants in the critical helix-helix interface segment, and find that efficient dimerization can be mediated by a wider range of sequence motifs than suggested by the earlier studies. We also show that certain mutants that are unable to dimerize can nevertheless form relatively high amounts of tetramers, and that specific tetramerization can be induced by duplication of the critical interface motif on the lipid-exposed side of the transmembrane helix.

Monomeric phospholamban overexpression in transgenic mouse hearts

Phospholamban, a prominent modulator of the sarcoplasmic reticulum (SR) Ca(2+)-ATPase activity and basal contractility in the mammalian heart, has been proposed to form pentamers in native SR membranes. However, the monomeric form of phospholamban, which is associated with mutating Cys41 to Phe41, was shown to be as effective as pentameric phospholamban in inhibiting Ca2+ transport in expression systems. To determine whether this monomeric form of phospholamban is also functional in vivo, we generated transgenic mice with cardiac-specific overexpression of the mutant (Cys41-->Phe41) phospholamban. Quantitative immunoblotting indicated a 2-fold increase in the cardiac phospholamban protein levels compared with wild-type controls, with approximately equal to 50% of phospholamban migrating as monomers and approximately 50% as pentamers upon SDS-PAGE. The mutant-phospholamban transgenic hearts were analyzed in parallel with transgenic hearts overexpressing (2-fold) wild-type phospholamban, which migrated as pentamers upon SDS-PAGE. SR Ca(2+)-uptake assays revealed that the EC50 values for Ca2+ were as follows: 0.32 +/- 0.01 mumol/L in hearts overexpressing monomeric phospholamban, 0.49 +/- 0.05 mumol/L in hearts overexpressing wild-type phospholamban, and 0.26 +/- 0.01 mumol/L in wild-type control mouse hearts. Analysis of cardiomyocyte mechanics and Ca2+ kinetics indicated that the inhibitory effects of mutant-phospholamban overexpression (mt) were less pronounced than those of wild-type phospholamban overexpression (ov) as assessed by depression of the following: (1) shortening fraction (25% mt versus 45% ov), (2) rates of shortening (27% mt versus 48% ov), (3) rates of relengthening (25% mt versus 50% ov) (4) amplitude of the Ca2+ signal (21% mt versus 40% ov), and (5) time for decay of the Ca2+ signal (25% mt versus 106% ov) compared with control (100%) myocytes. The differences in basal cardiac, myocyte mechanics and Ca2+ transients among the animal groups overexpressing monomeric or wild-type phospholamban and wild-type control mice were abolished upon isoproterenol stimulation. These findings suggest that pentameric assembly of phospholamban is important for mediating its optimal regulatory effects on myocardial contractility in vivo.

Purification of the cardiac sarcoplasmic reticulum membrane protein phospholamban from recombinant Escherichia coli

Phospholamban (PLN) was expressed in Escherichia coli as a protein fusion with glutathione S-transferase (GST). GST-PLN was mostly present in the insoluble protein fraction and accounted for approximately 50% of total insoluble protein. Attempts to suppress inclusion body formation or to use GST as an affinity-purification tag failed. A successful purification method is based on preparative SDS/PAGE and electrodialysis. From 1 g cells we typically purified 13.5 mg fusion protein with a PLN content of 2.8 mg. We genetically inserted an enterokinase (EK) protease site just in front of the PLN sequence and demonstrated the proteolytical liberation of PLN from the carrier protein. The approach described represents a substantial advancement in PLN expression and purification.

Keeping calcium in its place: Ca(2+)-ATPase and phospholamban

Electron microscopy is gradually revealing more and more about the structure of the calcium pump from the sarcoplasmic reticulum, Ca(2+)-ATPase. The most recent result reveals the ATP-binding site, and two different avenues are being pursued towards achieving a higher resolution structure. Although no such structures are currently available for phospholamban, various spectroscopies and site-directed mutagenesis have been combined to produce a compelling structural model for its regulation of Ca(2+)-ATPase.

Modeling transmembrane helical oligomers

Recently, methods for the analysis and design of water-soluble, oligomeric bundles of alpha helices, including coiled coils, have reached a high level of sophistication. These same methods may now be applied to transmembrane helical bundles. Studies of the transmembrane domains of glycophorin, phospholamban, and the M2 protein from influenza A virus exemplify this general approach.

Phospholamban inhibitory function is activated by depolymerization

Phospholamban (PLN), a homopentameric, integral membrane protein, reversibly inhibits cardiac sarcoplasmic reticulum Ca2+-ATPase (SERCA2a) activity through intramembrane interactions. Here, alanine-scanning mutagenesis of the PLN transmembrane sequence was used to identify two functional domains on opposite faces of the transmembrane helix. Mutations in one face diminish inhibitory interactions with transmembrane sequences of SERCA2a, but have relatively little effect on the pentameric state, while mutations in the other face activate inhibitory interactions and enhance monomer formation. Double mutants are monomeric, but loss of inhibitory function is dominant over activation of inhibitory function. These observations support the proposal that the SERCA2a interaction site lies on the helical face which is not involved in pentamer formation. Four highly inhibitory mutants are effectively devoid of pentamer, suggesting that pentameric PLN represents a less active or inactive reservoir that dissociates to provide inhibitory monomeric PLN subunits. A model is presented in which the degree of PLN inhibition of SERCA2a activity is ultimately determined by the concentration of the inhibited PLN monomer.SERCA2a heterodimeric complex. The concentration of this inhibited complex is determined by the dissociation constant for the PLN pentamer (which is mutation-sensitive) and by the dissociation constant for the PLN/SERCA2a heterodimer (which is likely to be mutation-sensitive).

Predicting coiled-coil regions in proteins

The past several years have seen significant advances in our ability to recognize coiled coils from protein sequences and model their structures. New methods include a detection program based on pairwise residue correlations, a program that distinguishes two-stranded from three-stranded coiled coils and a routine for modelling the coordinates of the core residues in coiled coils. Several widely noted predictions, among them those for heterotrimeric G proteins and for cartilage oligomeric matrix protein, have been confirmed by crystal structures, and several new predictions have been made, including a model for the still hypothetical right-handed coiled coil.

Alamethicin channels - modelling via restrained molecular dynamics simulations

Alamethicin channels have been modelled as approximately parallel bundles of transbilayer helices containing between N = 4 and 8 helices per bundle. Initial models were generated by in vacuo restrained molecular dynamics (MD) simulations, and were refined by 60 ps MD simulations with water molecules present within and at the mouths of the central pore. The helix bundles were stabilized by networks of H-bonds between intra-pore water molecules and Gln-7 side-chains. Channel conductances were predicted on the basis of pore radius profiles, and suggested that the N = 4 bundle formed an occluded pore, whereas pores with N > or = 5 helices per bundle were open. Continuum electrostatics calculations suggested that the N = 6 pore is cation-selective, whereas pores with N > or = 7 helices per bundle were predicted to be somewhat less ion-selective.

A novel method for structure-based prediction of ion channel conductance properties

A rapid and easy-to-use method of predicting the conductance of an ion channel from its three-dimensional structure is presented. The method combines the pore dimensions of the channel as measured in the HOLE program with an Ohmic model of conductance. An empirically based correction factor is then applied. The method yielded good results for six experimental channel structures (none of which were included in the training set) with predictions accurate to within an average factor of 1.62 to the true values. The predictive r2 was equal to 0.90, which is indicative of a good predictive ability. The procedure is used to validate model structures of alamethicin and phospholamban. Two genuine predictions for the conductance of channels with known structure but without reported conductances are given. A modification of the procedure that calculates the expected results for the effect of the addition of nonelectrolyte polymers on conductance is set out. Results for a cholera toxin B-subunit crystal structure agree well with the measured values. The difficulty in interpreting such studies is discussed, with the conclusion that measurements on channels of known structure are required.

Structural perspectives of phospholamban, a helical transmembrane pentamer

Phospholamban is a 52-amino-acid protein that assembles into a pentamer in sarcoplasmic reticulum membranes. The protein has a role in the regulation of the resident calcium ATPase through an inhibitory association that can be reversed by phosphorylation. The phosphorylation of phospholamban is initiated by beta-adrenergic stimulation, identifying phospholamban as an important component in the stimulation of cardiac activity by beta-agonists. In this role of phospholamban that has motivated studies in recent decades. There is evidence that phospholamban may also function as a Ca(2+)-selective ion channel. The structural properties of phospholamban have been studied by mutagenesis, modeling, and spectroscopy, resulting in a new view of the organization of this key molecule in membranes.

HOLE: a program for the analysis of the pore dimensions of ion channel structural models

A method (HOLE) that allows the analysis of the dimensions of the pore running through a structural model of an ion channel is presented. The algorithm uses a Monte Carlo simulated annealing procedure to find the best route for a sphere with variable radius to squeeze through the channel. Results can be displayed in a graphical fashion or visualized with most common molecular graphical packages. Advances include a method to analyze the anisotropy within a pore. The method can also be used to predict the conductance of channels using a simple empirically corrected ohmic model. As an example the program is applied to the cholera toxin B-subunit pentamer. The compatibility of the crystal structure and conductance data is established.

Improved prediction for the structure of the dimeric transmembrane domain of glycophorin A obtained through global searching

A more global search method, using fewer assumptions, has been used to predict the structure of the dimeric transmembrane region of the protein glycophorin A. The resulting model significantly differs from that previously determined. In particular, the arrangement between the two transmembrane helices is now more symmetric resulting in improved interaction energies and an increased buried surface area. An increase in the van der Waals interaction energy due to tighter packing compensates for the loss of the interhelical hydrogen bond observed between Thr-87 of each helix in the previous model.

The crystal structure of a five-stranded coiled coil in COMP: a prototype ion channel?

Oligomerization by the formation of alpha-helical bundles is common in many proteins. The crystal structure of a parallel pentameric coiled coil, constituting the oligomerization domain in the cartilage oligomeric matrix protein (COMP), was determined at 2.05 angstroms resolution. The same structure probably occurs in two other extracellular matrix proteins, thrombospondins 3 and 4. Complementary hydrophobic interactions and conserved disulfide bridges between the alpha helices result in a thermostable structure with unusual properties. The long hydrophobic axial pore is filled with water molecules but can also accommodate small apolar groups. An "ion trap" is formed inside the pore by a ring of conserved glutamines, which binds chloride and probably other monatomic anions. The oligomerization domain of COMP has marked similarities with proposed models of the pentameric transmembrane ion channels in phospholamban and the acetylcholine receptor.