Evolutionary-guided de novo structure prediction of self-associated transmembrane helical proteins with near-atomic accuracy

TMKit: a Python interface for computational analysis of transmembrane proteins

Transmembrane proteins are receptors, enzymes, transporters and ion channels that are instrumental in regulating a variety of cellular activities, such as signal transduction and cell communication. Despite tremendous progress in computational capacities to support protein research, there is still a significant gap in the availability of specialized computational analysis toolkits for transmembrane protein research. Here, we introduce TMKit, an open-source Python programming interface that is modular, scalable and specifically designed for processing transmembrane protein data. TMKit is a one-stop computational analysis tool for transmembrane proteins, enabling users to perform database wrangling, engineer features at the mutational, domain and topological levels, and visualize protein-protein interaction interfaces. In addition, TMKit includes seqNetRR, a high-performance computing library that allows customized construction of a large number of residue connections. This library is particularly well suited for assigning correlation matrix-based features at a fast speed. TMKit should serve as a useful tool for researchers in assisting the study of transmembrane protein sequences and structures. TMKit is publicly available through https://github.com/2003100127/tmkit and https://tmkit-guide.herokuapp.com/doc/overview.

Deep transfer learning for inter-chain contact predictions of transmembrane protein complexes

Membrane proteins are encoded by approximately a quarter of human genes. Inter-chain residue-residue contact information is important for structure prediction of membrane protein complexes and valuable for understanding their molecular mechanism. Although many deep learning methods have been proposed to predict the intra-protein contacts or helix-helix interactions in membrane proteins, it is still challenging to accurately predict their inter-chain contacts due to the limited number of transmembrane proteins. Addressing the challenge, here we develop a deep transfer learning method for predicting inter-chain contacts of transmembrane protein complexes, named DeepTMP, by taking advantage of the knowledge pre-trained from a large data set of non-transmembrane proteins. DeepTMP utilizes a geometric triangle-aware module to capture the correct inter-chain interaction from the coevolution information generated by protein language models. DeepTMP is extensively evaluated on a test set of 52 self-associated transmembrane protein complexes, and compared with state-of-the-art methods including DeepHomo2.0, CDPred, GLINTER, DeepHomo, and DNCON2_Inter. It is shown that DeepTMP considerably improves the precision of inter-chain contact prediction and outperforms the existing approaches in both accuracy and robustness.

Computationally designed GPCR quaternary structures bias signaling pathway activation

Communication across membranes controls critical cellular processes and is achieved by receptors translating extracellular signals into selective cytoplasmic responses. While receptor tertiary structures can be readily characterized, receptor associations into quaternary structures are challenging to study and their implications in signal transduction remain poorly understood. Here, we report a computational approach for predicting receptor self-associations, and designing receptor oligomers with various quaternary structures and signaling properties. Using this approach, we designed chemokine receptor CXCR4 dimers with reprogrammed binding interactions, conformations, and abilities to activate distinct intracellular signaling proteins. In agreement with our predictions, the designed CXCR4s dimerized through distinct conformations and displayed different quaternary structural changes upon activation. Consistent with the active state models, all engineered CXCR4 oligomers activated the G protein Gi, but only specific dimer structures also recruited β-arrestins. Overall, we demonstrate that quaternary structures represent an important unforeseen mechanism of receptor biased signaling and reveal the existence of a bias switch at the dimer interface of several G protein-coupled receptors including CXCR4, mu-Opioid and type-2 Vasopressin receptors that selectively control the activation of G proteins vs β-arrestin-mediated pathways. The approach should prove useful for predicting and designing receptor associations to uncover and reprogram selective cellular signaling functions.

Molecular mechanism of interaction between fatty acid delta 6 desaturase and acyl-CoA by computational prediction

Enzyme catalyzed desaturation of intracellular fatty acids plays an important role in various physiological and pathological processes related to lipids. Limited to the multiple transmembrane domains, it is difficult to obtain their three-dimensional structure of fatty acid desaturases. So how they interact with their substrates is unclear. Here, we predicted the complex of Micromonas pusilla delta 6 desaturase (MpFADS6) with the substrate linoleinyl-CoA (ALA-CoA) by trRosetta software and docking poses by Dock 6 software. The potential enzyme–substrate binding sites were anchored by analysis of the complex. Then, site-directed mutagenesis and activity verification clarified that W290, W224, and F352 were critical residues of the substrate tunnel and directly bonded to ALA-CoA. H94 and H69 were indispensable for transporting electrons with heme. H452, N445, and H358 significantly influenced the recognition and attraction of MpFADS6 to the substrate. These findings provide new insights and methods to determine the structure, mechanisms and directed transformation of membrane-bound desaturases.

The structure of the Δ6 fatty acid desaturase and substrate complex is modeled.

The substrate tunnel and key residues of MpFADS6 catalytic activity are determined.

The new insights to determine the mechanism of the membrane-bound desaturases.

Deep learning approaches for conformational flexibility and switching properties in protein design

Following the hugely successful application of deep learning methods to protein structure prediction, an increasing number of design methods seek to leverage generative models to design proteins with improved functionality over native proteins or novel structure and function. The inherent flexibility of proteins, from side-chain motion to larger conformational reshuffling, poses a challenge to design methods, where the ideal approach must consider both the spatial and temporal evolution of proteins in the context of their functional capacity. In this review, we highlight existing methods for protein design before discussing how methods at the forefront of deep learning-based design accommodate flexibility and where the field could evolve in the future.

Coarse Grained Molecular Dynamic Simulations for the Study of TNF Receptor Family Members' Transmembrane Organization

The Tumor Necrosis Factor (TNF) and the TNF receptor (TNFR) superfamilies are composed of 19 ligands and 30 receptors, respectively. The oligomeric properties of ligands, both membrane bound and soluble, has been studied most. However, less is known about the oligomeric properties of TNFRs. Earlier reports identified the extracellular, membrane-distal, cysteine-rich domain as a pre-ligand assembly domain which stabilizes receptor dimers and/or trimers in the absence of ligand. Nevertheless, recent reports based on structural nuclear magnetic resonance (NMR) highlight the intrinsic role of the transmembrane domains to form dimers (p75NTR), trimers (Fas), or dimers of trimers (DR5). Thus, understanding the structural basis of transmembrane oligomerization may shed light on the mechanism for signal transduction and the impact of disease-associated mutations in this region. To this end, here we used an in silico coarse grained molecular dynamics approach with Martini force field to study TNFR transmembrane homotypic interactions. We have first validated this approach studying the three TNFR described by NMR (p75NTR, Fas, and DR5). We have simulated membrane patches composed of 36 helices of the same receptor equidistantly distributed in order to get unbiassed information on spontaneous proteins assemblies. Good agreement was found in the specific residues involved in homotypic interactions and we were able to observe dimers, trimers, and higher-order oligomers corresponding to those reported in NMR experiments. We have, applied this approach to study the assembly of disease-related mutations being able to assess their impact on oligomerization stability. In conclusion, our results showed the usefulness of coarse grained simulations with Martini force field to study in an unbiased manner higher order transmembrane oligomerization.

Experimental determination and data-driven prediction of homotypic transmembrane domain interfaces

•

Homotypic TMD interfaces identified by different techniques share strong similarities.

•

The GxxxG motif is the feature most strongly associated with interfaces.

•

Other features include conservation, polarity, coevolution, and depth in the membrane

•

The role of each of each feature strongly depends on the individual protein.

•

Machine-learning helps predict interfaces from evolutionary sequence data

Interactions between their transmembrane domains (TMDs) frequently support the assembly of single-pass membrane proteins to non-covalent complexes. Yet, the TMD-TMD interactome remains largely uncharted. With a view to predicting homotypic TMD-TMD interfaces from primary structure, we performed a systematic analysis of their physical and evolutionary properties. To this end, we generated a dataset of 50 self-interacting TMDs. This dataset contains interfaces of nine TMDs from bitopic human proteins (Ire1, Armcx6, Tie1, ATP1B1, PTPRO, PTPRU, PTPRG, DDR1, and Siglec7) that were experimentally identified here and combined with literature data. We show that interfacial residues of these homotypic TMD-TMD interfaces tend to be more conserved, coevolved and polar than non-interfacial residues. Further, we suggest for the first time that interface positions are deficient in β-branched residues, and likely to be located deep in the hydrophobic core of the membrane. Overrepresentation of the GxxxG motif at interfaces is strong, but that of (small)xxx(small) motifs is weak. The multiplicity of these features and the individual character of TMD-TMD interfaces, as uncovered here, prompted us to train a machine learning algorithm. The resulting prediction method, THOIPA (www.thoipa.org), excels in the prediction of key interface residues from evolutionary sequence data.

Refinement of evolutionary medicine predictions based on clinical evidence for the manifestations of Mendelian diseases

Prediction methods have become an integral part of biomedical and biotechnological research. However, their clinical interpretations are largely based on biochemical or molecular data, but not clinical data. Here, we focus on improving the reliability and clinical applicability of prediction algorithms. We assembled and curated two large non-overlapping large databases of clinical phenotypes. These phenotypes were caused by missense variations in 44 and 63 genes associated with Mendelian diseases. We used these databases to establish and validate the model, allowing us to improve the predictions obtained from EVmutation, SNAP2 and PoPMuSiC 2.1. The predictions of clinical effects suffered from a lack of specificity, which appears to be the common constraint of all recently used prediction methods, although predictions mediated by these methods are associated with nearly absolute sensitivity. We introduced evidence-based tailoring of the default settings of the prediction methods; this tailoring substantially improved the prediction outcomes. Additionally, the comparisons of the clinically observed and theoretical variations led to the identification of large previously unreported pools of variations that were under negative selection during molecular evolution. The evolutionary variation analysis approach described here is the first to enable the highly specific identification of likely disease-causing missense variations that have not yet been associated with any clinical phenotype.

Computational Nanoscopy of Tight Junctions at the Blood-Brain Barrier Interface

The selectivity of the blood-brain barrier (BBB) is primarily maintained by tight junctions (TJs), which act as gatekeepers of the paracellular space by blocking blood-borne toxins, drugs, and pathogens from entering the brain. The BBB presents a significant challenge in designing neurotherapeutics, so a comprehensive understanding of the TJ architecture can aid in the design of novel therapeutics. Unraveling the intricacies of TJs with conventional experimental techniques alone is challenging, but recently developed computational tools can provide a valuable molecular-level understanding of TJ architecture. We employed the computational methods toolkit to investigate claudin-5, a highly expressed TJ protein at the BBB interface. Our approach started with the prediction of claudin-5 structure, evaluation of stable dimer conformations and nanoscale assemblies, followed by the impact of lipid environments, and posttranslational modifications on these claudin-5 assemblies. These led to the study of TJ pores and barriers and finally understanding of ion and small molecule transport through the TJs. Some of these in silico, molecular-level findings, will need to be corroborated by future experiments. The resulting understanding can be advantageous towards the eventual goal of drug delivery across the BBB. This review provides key insights gleaned from a series of state-of-the-art nanoscale simulations (or computational nanoscopy studies) performed on the TJ architecture.

Bio-knowledge-based filters improve residue-residue contact prediction accuracy

Motivation

Residue-residue contact prediction through direct coupling analysis has reached impressive accuracy, but yet higher accuracy will be needed to allow for routine modelling of protein structures. One way to improve the prediction accuracy is to filter predicted contacts using knowledge about the particular protein of interest or knowledge about protein structures in general.

Results

We focus on the latter and discuss a set of filters that can be used to remove false positive contact predictions. Each filter depends on one or a few cut-off parameters for which the filter performance was investigated. Combining all filters while using default parameters resulted for a test set of 851 protein domains in the removal of 29% of the predictions of which 92% were indeed false positives.

Availability and implementation

All data and scripts are available at http://comprec-lin.iiar.pwr.edu.pl/FPfilter/.

Supplementary information

Supplementary data are available at Bioinformatics online.

A lipophilicity-based energy function for membrane-protein modelling and design

Membrane-protein design is an exciting and increasingly successful research area which has led to landmarks including the design of stable and accurate membrane-integral proteins based on coiled-coil motifs. Design of topologically more complex proteins, such as most receptors, channels, and transporters, however, demands an energy function that balances contributions from intra-protein contacts and protein-membrane interactions. Recent advances in water-soluble all-atom energy functions have increased the accuracy in structure-prediction benchmarks. The plasma membrane, however, imposes different physical constraints on protein solvation. To understand these constraints, we recently developed a high-throughput experimental screen, called dsTβL, and inferred apparent insertion energies for each amino acid at dozens of positions across the bacterial plasma membrane. Here, we express these profiles as lipophilicity energy terms in Rosetta and demonstrate that the new energy function outperforms previous ones in modelling and design benchmarks. Rosetta ab initio simulations starting from an extended chain recapitulate two-thirds of the experimentally determined structures of membrane-spanning homo-oligomers with <2.5Å root-mean-square deviation within the top-predicted five models (available online: http://tmhop.weizmann.ac.il). Furthermore, in two sequence-design benchmarks, the energy function improves discrimination of stabilizing point mutations and recapitulates natural membrane-protein sequences of known structure, thereby recommending this new energy function for membrane-protein modelling and design.

Structural Model of the ETR1 Ethylene Receptor Transmembrane Sensor Domain

The structure, mechanism of action and copper stoichiometry of the transmembrane sensor domain of the plant ethylene receptor ETR1 and homologs have remained elusive, hampering the understanding on how the perception of the plant hormone ethylene is transformed into a downstream signal. We generated the first structural model of the transmembrane sensor domain of ETR1 by integrating ab initio structure prediction and coevolutionary information. To refine and independently validate the model, we determined protein-related copper stoichiometries on purified receptor preparations and explored the helix arrangement by tryptophan scanning mutagenesis. All-atom molecular dynamics simulations of the dimeric model reveal how ethylene can bind proximal to the copper ions in the receptor, illustrating the initial stages of the ethylene perception process.

Interfaces Between Alpha-helical Integral Membrane Proteins: Characterization, Prediction, and Docking

Protein-protein interaction (PPI) is an essential mechanism by which proteins perform their biological functions. For globular proteins, the molecular characteristics of such interactions have been well analyzed, and many computational tools are available for predicting PPI sites and constructing structural models of the complex. In contrast, little is known about the molecular features of the interaction between integral membrane proteins (IMPs) and few methods exist for constructing structural models of their complexes. Here, we analyze the interfaces from a non-redundant set of complexes of α-helical IMPs whose structures have been determined to a high resolution. We find that the interface is not significantly different from the rest of the surface in terms of average hydrophobicity. However, the interface is significantly better conserved and, on average, inter-subunit contacting residue pairs correlate more strongly than non-contacting pairs, especially in obligate complexes. We also develop a neural network-based method, with an area under the receiver operating characteristic curve of 0.75 and a Pearson correlation coefficient of 0.70, for predicting interface residues and their weighted contact numbers (WCNs). We further show that predicted interface residues and their WCNs can be used as restraints to reconstruct the structure α-helical IMP dimers through docking for fourteen out of a benchmark set of sixteen complexes. The RMSD100 values of the best-docked ligand subunit to its native structure are <2.5 Å for these fourteen cases. The structural analysis conducted in this work provides molecular details about the interface between α-helical IMPs and the WCN restraints represent an efficient means to score α-helical IMP docking candidates.

De novo designed transmembrane peptides activating the α5β1 integrin

Computationally designed transmembrane α-helical peptides (CHAMP) have been used to compete for helix-helix interactions within the membrane, enabling the ability to probe the activation of the integrins αIIbβ3 and αvβ3. Here, this method is extended towards the design of CHAMP peptides that inhibit the association of the α5β1 transmembrane (TM) domains, targeting the Ala-X3-Gly motif within α5. Our previous design algorithm was performed alongside a new workflow implemented within the widely used Rosetta molecular modeling suite. Peptides from each computational approach activated integrin α5β1 but not αVβ3 in human endothelial cells. Two CHAMP peptides were shown to directly associate with an α5 TM domain peptide in detergent micelles to a similar degree as a β1 TM peptide does. By solution-state nuclear magnetic resonance, one of these CHAMP peptides was shown to bind primarily the integrin β1 TM domain, which itself has a Gly-X3-Gly motif. The second peptide associated modestly with both α5 and β1 constructs, with slight preference for α5. Although the design goal was not fully realized, this work characterizes novel CHAMP peptides activating α5β1 that can serve as useful reagents for probing integrin biology.

NMR relaxation parameters of methyl groups as a tool to map the interfaces of helix-helix interactions in membrane proteins

In the case of soluble proteins, chemical shift mapping is used to identify the intermolecular interfaces when the NOE-based calculations of spatial structure of the molecular assembly are impossible or impracticable. However, the reliability of the membrane protein interface mapping based on chemical shifts or other relevant parameters was never assessed. In the present work, we investigate the predictive power of various NMR parameters that can be used for mapping of helix-helix interfaces in dimeric TM domains. These parameters are studied on a dataset containing three structures of helical dimers obtained for two different proteins in various membrane mimetics. We conclude that the amide chemical shifts have very little predictive value, while the methyl chemical shifts could be used to predict interfaces, though with great care. We suggest an approach based on conversion of the carbon NMR relaxation parameters of methyl groups into parameters of motion, and one of such values, the characteristic time of methyl rotation, appears to be a reliable sensor of interhelix contacts in transmembrane domains. The carbon NMR relaxation parameters of methyl groups can be measured accurately and with high sensitivity and resolution, making the proposed parameter a useful tool for investigation of protein-protein interfaces even in large membrane proteins. An approach to build the models of transmembrane dimers based on perturbations of methyl parameters and TMDOCK software is suggested.

Forecasting residue-residue contact prediction accuracy

Motivation

Apart from meta-predictors, most of today's methods for residue-residue contact prediction are based entirely on Direct Coupling Analysis (DCA) of correlated mutations in multiple sequence alignments (MSAs). These methods are on average ∼40% correct for the 100 strongest predicted contacts in each protein. The end-user who works on a single protein of interest will not know if predictions are either much more or much less correct than 40%, which is especially a problem if contacts are predicted to steer experimental research on that protein.

Results

We designed a regression model that forecasts the accuracy of residue-residue contact prediction for individual proteins with an average error of 7 percentage points. Contacts were predicted with two DCA methods (gplmDCA and PSICOV). The models were built on parameters that describe the MSA, the predicted secondary structure, the predicted solvent accessibility and the contact prediction scores for the target protein. Results show that our models can be also applied to the meta-methods, which was tested on RaptorX.

Availability and implementation

All data and scripts are available from http://comprec-lin.iiar.pwr.edu.pl/dcaQ/.

Contact

malgorzata.kotulska@pwr.edu.pl.

Supplementary information

Supplementary data are available at Bioinformatics online.

Enhanced unbiased sampling of protein dynamics using evolutionary coupling information

One of the major challenges in atomistic simulations of proteins is efficient sampling of pathways associated with rare conformational transitions. Recent developments in statistical methods for computation of direct evolutionary couplings between amino acids within and across polypeptide chains have allowed for inference of native residue contacts, informing accurate prediction of protein folds and multimeric structures. In this study, we assess the use of distances between evolutionarily coupled residues as natural choices for reaction coordinates which can be incorporated into Markov state model-based adaptive sampling schemes and potentially used to predict not only functional conformations but also pathways of conformational change, protein folding, and protein-protein association. We demonstrate the utility of evolutionary couplings in sampling and predicting activation pathways of the β 2-adrenergic receptor (β 2-AR), folding of the FiP35 WW domain, and dimerization of the E. coli molybdopterin synthase subunits. We find that the time required for β 2-AR activation and folding of the WW domain are greatly diminished using evolutionary couplings-guided adaptive sampling. Additionally, we were able to identify putative molybdopterin synthase association pathways and near-crystal structure complexes from protein-protein association simulations.

Rosetta Broker for membrane protein structure prediction: concentrative nucleoside transporter 3 and corticotropin-releasing factor receptor 1 test cases

Background

Membrane proteins are difficult targets for structure prediction due to the limited structural data deposited in Protein Data Bank. Most computational methods for membrane protein structure prediction are based on the comparative modeling. There are only few de novo methods targeting that distinct protein family. In this work an example of such de novo method was used to structurally and functionally characterize two representatives of distinct membrane proteins families of solute carrier transporters and G protein-coupled receptors. The well-known Rosetta program and one of its protocols named Broker was used in two test cases. The first case was de novo structure prediction of three N-terminal transmembrane helices of the human concentrative nucleoside transporter 3 (hCNT3) homotrimer belonging to the solute carrier 28 family of transporters (SLC28). The second case concerned the large scale refinement of transmembrane helices of a homology model of the corticotropin-releasing factor receptor 1 (CRFR1) belonging to the G protein-coupled receptors family.

Results

The inward-facing model of the hCNT3 homotrimer was used to propose the functional impact of its single nucleotide polymorphisms. Additionally, the 100 ns molecular dynamics simulation of the unliganded hCNT3 model confirmed its validity and revealed mobility of the selected binding site and homotrimer interface residues. The large scale refinement of transmembrane helices of the CRFR1 homology model resulted in the significant improvement of its accuracy with respect to the crystal structure of CRFR1, especially in the binding site area. Consequently, the antagonist CP-376395 could be docked with Autodock VINA to the CRFR1 model without any steric clashes.

Conclusions

The presented work demonstrated that Rosetta Broker can be a versatile tool for solving various issues referring to protein biology. Two distinct examples of de novo membrane protein structure prediction presented here provided important insights into three major areas of protein biology. Namely, the dynamics of the inward-facing hCNT3 homotrimer system, the structural changes of the CRFR1 receptor upon the antagonist binding and finally, the role of single nucleotide polymorphisms in both, hCNT3 and CRFR1 proteins, were investigated.

Electronic supplementary material

The online version of this article (doi:10.1186/s12900-017-0078-8) contains supplementary material, which is available to authorized users.

The Rosetta All-Atom Energy Function for Macromolecular Modeling and Design

Over the past decade, the Rosetta biomolecular modeling suite has informed diverse biological questions and engineering challenges ranging from interpretation of low-resolution structural data to design of nanomaterials, protein therapeutics, and vaccines. Central to Rosetta's success is the energy function: a model parametrized from small-molecule and X-ray crystal structure data used to approximate the energy associated with each biomolecule conformation. This paper describes the mathematical models and physical concepts that underlie the latest Rosetta energy function, called the Rosetta Energy Function 2015 (REF15). Applying these concepts, we explain how to use Rosetta energies to identify and analyze the features of biomolecular models. Finally, we discuss the latest advances in the energy function that extend its capabilities from soluble proteins to also include membrane proteins, peptides containing noncanonical amino acids, small molecules, carbohydrates, nucleic acids, and other macromolecules.

Toward high-resolution computational design of the structure and function of helical membrane proteins

The computational design of α-helical membrane proteins is still in its infancy but has already made great progress. De novo design allows stable, specific and active minimal oligomeric systems to be obtained. Computational reengineering can improve the stability and function of naturally occurring membrane proteins. Currently, the major hurdle for the field is the experimental characterization of the designs. The emergence of new structural methods for membrane proteins will accelerate progress.

Applications of contact predictions to structural biology

Evolutionary pressure on residue interactions, intramolecular or intermolecular, that are important for protein structure or function can lead to covariance between the two positions. Recent methodological advances allow much more accurate contact predictions to be derived from this evolutionary covariance signal. The practical application of contact predictions has largely been confined to structural bioinformatics, yet, as this work seeks to demonstrate, the data can be of enormous value to the structural biologist working in X-ray crystallo-graphy, cryo-EM or NMR. Integrative structural bioinformatics packages such as Rosetta can already exploit contact predictions in a variety of ways. The contribution of contact predictions begins at construct design, where structural domains may need to be expressed separately and contact predictions can help to predict domain limits. Structure solution by molecular replacement (MR) benefits from contact predictions in diverse ways: in difficult cases, more accurate search models can be constructed using ab initio modelling when predictions are available, while intermolecular contact predictions can allow the construction of larger, oligomeric search models. Furthermore, MR using supersecondary motifs or large-scale screens against the PDB can exploit information, such as the parallel or antiparallel nature of any β-strand pairing in the target, that can be inferred from contact predictions. Contact information will be particularly valuable in the determination of lower resolution structures by helping to assign sequence register. In large complexes, contact information may allow the identity of a protein responsible for a certain region of density to be determined and then assist in the orientation of an available model within that density. In NMR, predicted contacts can provide long-range information to extend the upper size limit of the technique in a manner analogous but complementary to experimental methods. Finally, predicted contacts can distinguish between biologically relevant interfaces and mere lattice contacts in a final crystal structure, and have potential in the identification of functionally important regions and in foreseeing the consequences of mutations.

Membranome: a database for proteome-wide analysis of single-pass membrane proteins

The Membranome database was developed to assist analysis and computational modeling of single-pass (bitopic) transmembrane (TM) proteins and their complexes by providing structural information about these proteins on a genomic scale. The database currently collects data on >6000 bitopic proteins from Homo sapiens, Arabidopsis thaliana, Dictyostelium discoideum, Saccharomyces cerevisiae, Escherichia coli and Methanocaldococcus jannaschii It presents the following data: (i) hierarchical classification of bitopic proteins into 15 functional classes, 689 structural superfamilies and 1404 families; (ii) 446 complexes of bitopic proteins with known three-dimensional (3D) structures classified into 129 families; (iii) computationally generated three-dimensional models of TM α-helices positioned in membranes; (iv) amino acid sequences, domain architecture, functional annotation and available experimental structures of bitopic proteins; (v) TM topology and intracellular localization, (vi) physical interactions between proteins from the database along with links to other resources. The database is freely accessible at http://membranome.org There is a variety of options for browsing, sorting, searching and retrieval of the content, including downloadable coordinate files of TM domains with calculated membrane boundaries.

Assembling the Tat protein translocase

The twin-arginine protein translocation system (Tat) transports folded proteins across the bacterial cytoplasmic membrane and the thylakoid membranes of plant chloroplasts. The Tat transporter is assembled from multiple copies of the membrane proteins TatA, TatB, and TatC. We combine sequence co-evolution analysis, molecular simulations, and experimentation to define the interactions between the Tat proteins of Escherichia coli at molecular-level resolution. In the TatBC receptor complex the transmembrane helix of each TatB molecule is sandwiched between two TatC molecules, with one of the inter-subunit interfaces incorporating a functionally important cluster of interacting polar residues. Unexpectedly, we find that TatA also associates with TatC at the polar cluster site. Our data provide a structural model for assembly of the active Tat translocase in which substrate binding triggers replacement of TatB by TatA at the polar cluster site. Our work demonstrates the power of co-evolution analysis to predict protein interfaces in multi-subunit complexes.

DOI:http://dx.doi.org/10.7554/eLife.20718.001

Drugging Membrane Protein Interactions

The majority of therapeutics target membrane proteins, accessible on the surface of cells, to alter cellular signaling. Cells use membrane proteins to transduce signals into cells, transport ions and molecules, bind cells to a surface or substrate, and catalyze reactions. Newly devised technologies allow us to drug conventionally "undruggable" regions of membrane proteins, enabling modulation of protein-protein, protein-lipid, and protein-nucleic acid interactions. In this review, we survey the state of the art of high-throughput screening and rational design in drug discovery, and we evaluate the advances in biological understanding and technological capacity that will drive pharmacotherapy forward against unorthodox membrane protein targets.

Sequence co-evolutionary information is a natural partner to minimally-frustrated models of biomolecular dynamics

Experimentally derived structural constraints have been crucial to the implementation of computational models of biomolecular dynamics. For example, not only does crystallography provide essential starting points for molecular simulations but also high-resolution structures permit for parameterization of simplified models. Since the energy landscapes for proteins and other biomolecules have been shown to be minimally frustrated and therefore funneled, these structure-based models have played a major role in understanding the mechanisms governing folding and many functions of these systems. Structural information, however, may be limited in many interesting cases. Recently, the statistical analysis of residue co-evolution in families of protein sequences has provided a complementary method of discovering residue-residue contact interactions involved in functional configurations. These functional configurations are often transient and difficult to capture experimentally. Thus, co-evolutionary information can be merged with that available for experimentally characterized low free-energy structures, in order to more fully capture the true underlying biomolecular energy landscape.

A topological and conformational stability alphabet for multipass membrane proteins

Multipass membrane proteins perform critical signal transduction and transport across membranes. How transmembrane helix (TMH) sequences encode the topology and conformational flexibility regulating these functions remains poorly understood. Here we describe a comprehensive analysis of the sequence-structure relationships at multiple interacting TMHs from all membrane proteins with structures in the Protein Data Bank (PDB). We found that membrane proteins can be deconstructed in interacting TMH trimer units, which mostly fold into six distinct structural classes of topologies and conformations. Each class is enriched in recurrent sequence motifs from functionally unrelated proteins, revealing unforeseen consensus and evolutionary conserved networks of stabilizing interhelical contacts. Interacting TMHs' topology and local protein conformational flexibility were remarkably well predicted in a blinded fashion from the identified binding-hotspot motifs. Our results reveal universal sequence-structure principles governing the complex anatomy and plasticity of multipass membrane proteins that may guide de novo structure prediction, design, and studies of folding and dynamics.

An Integrated Framework Advancing Membrane Protein Modeling and Design

Membrane proteins are critical functional molecules in the human body, constituting more than 30% of open reading frames in the human genome. Unfortunately, a myriad of difficulties in overexpression and reconstitution into membrane mimetics severely limit our ability to determine their structures. Computational tools are therefore instrumental to membrane protein structure prediction, consequently increasing our understanding of membrane protein function and their role in disease. Here, we describe a general framework facilitating membrane protein modeling and design that combines the scientific principles for membrane protein modeling with the flexible software architecture of Rosetta3. This new framework, called RosettaMP, provides a general membrane representation that interfaces with scoring, conformational sampling, and mutation routines that can be easily combined to create new protocols. To demonstrate the capabilities of this implementation, we developed four proof-of-concept applications for (1) prediction of free energy changes upon mutation; (2) high-resolution structural refinement; (3) protein-protein docking; and (4) assembly of symmetric protein complexes, all in the membrane environment. Preliminary data show that these algorithms can produce meaningful scores and structures. The data also suggest needed improvements to both sampling routines and score functions. Importantly, the applications collectively demonstrate the potential of combining the flexible nature of RosettaMP with the power of Rosetta algorithms to facilitate membrane protein modeling and design.

Role of GxxxG Motifs in Transmembrane Domain Interactions

Transmembrane (TM) helices of integral membrane proteins can facilitate strong and specific noncovalent protein-protein interactions. Mutagenesis and structural analyses have revealed numerous examples in which the interaction between TM helices of single-pass membrane proteins is dependent on a GxxxG or (small)xxx(small) motif. It is therefore tempting to use the presence of these simple motifs as an indicator of TM helix interactions. In this Current Topic review, we point out that these motifs are quite common, with more than 50% of single-pass TM domains containing a (small)xxx(small) motif. However, the actual interaction strength of motif-containing helices depends strongly on sequence context and membrane properties. In addition, recent studies have revealed several GxxxG-containing TM domains that interact via alternative interfaces involving hydrophobic, polar, aromatic, or even ionizable residues that do not form recognizable motifs. In multipass membrane proteins, GxxxG motifs can be important for protein folding, and not just oligomerization. Our current knowledge thus suggests that the presence of a GxxxG motif alone is a weak predictor of protein dimerization in the membrane.