A Systematic Search Method for the Identification of Tightly Packed Transmembrane Parallel α-Helices

Abstract

Membrane proteins play a major role in number of biological processes such as signaling pathways. The determination of the three-dimensional structure of these proteins is increasingly important for our understanding of their structure-function relationships. Due to the difficulty in isolating membrane proteins for X-ray diffraction studies, computational techniques are being developed to generate the 3D structures of TM domains. Here, we present a systematic search method for the identification of energetically favorable and tightly packed transmembrane parallel α-helices. The first step in our systematic search method is the generation of 3D models for pairs of parallel helix bundles with all possible orientations followed by an energy-based filter to eliminate structures with severe non-bonded contacts. Then, a RMS-based filter was used to cluster these structures into families. Furthermore, these dimers were energy minimized using molecular mechanics force field. Finally, we identified the tightly packed parallel α-helices by using an interface surface area. To validate our search method, we compared our predicted GlycophorinA dimer structures with the reported NMR structures. With our search method, we are able to reproduce NMR structures of GPA with 0.9Å RMSD. In addition, by considering the reported mutational data on GxxxG motif interactions, twenty percent of our predicted dimers are within in the 2.0Å RMSD. The dimers obtained from our method were used to generate parallel trimeric and tetramer TM structures of GPA and found that the structure of GPA might exist only in a dimer form as reported earlier.     

A novel scoring function for predicting the conformations of tightly packed pairs of transmembrane alpha-helices

Pairs of helices in transmembrane (TM) proteins are often tightly packed. We present a scoring function and a computational methodology for predicting the tertiary fold of a pair of alpha-helices such that its chances of being tightly packed are maximized. Since the number of TM protein structures solved to date is small, it seems unlikely that a reliable scoring function derived statistically from the known set of TM protein structures will be available in the near future. We therefore constructed a scoring function based on the qualitative insights gained in the past two decades from the solved structures of TM and soluble proteins. In brief, we reward the formation of contacts between small amino acid residues such as Gly, Cys, and Ser, that are known to promote dimerization of helices, and penalize the burial of large amino acid residues such as Arg and Trp. As a case study, we show that our method predicts the native structure of the TM homodimer glycophorin A (GpA) to be, in essence, at the global score optimum. In addition, by correlating our results with empirical point mutations on this homodimer, we demonstrate that our method can be a helpful adjunct to mutation analysis. We present a data set of canonical alpha-helices from the solved structures of TM proteins and provide a set of programs for analyzing it (http://ashtoret.tau.ac.il/~sarel). From this data set we derived 11 helix pairs, and conducted searches around their native states as a further test of our method. Approximately 73% of our predictions showed a reasonable fit (RMS deviation <2A) with the native structures compared to the success rate of 8% expected by chance. The search method we employ is less effective for helix pairs that are connected via short loops (<20 amino acid residues), indicating that short loops may play an important role in determining the conformation of alpha-helices in TM proteins.     

Modeling Transmembrane Domain Dimers/Trimers of Plexin Receptors: Implications for Mechanisms of Signal Transmission across the Membrane

Single-pass transmembrane (TM) receptors transmit signals across lipid bilayers by helix association or by configurational changes within preformed dimers. The structure determination for such TM regions is challenging and has mostly been accomplished by NMR spectroscopy. Recently, the computational prediction of TM dimer structures is becoming recognized for providing models, including alternate conformational states, which are important for receptor regulation. Here we pursued a strategy to predict helix oligomers that is based on packing considerations (using the PREDDIMER webserver) and is followed by a refinement of structures, utilizing microsecond all-atom molecular dynamics simulations. We applied this method to plexin TM receptors, a family of 9 human proteins, involved in the regulation of cell guidance and motility. The predicted models show that, overall, the preferences identified by PREDDIMER are preserved in the unrestrained simulations and that TM structures are likely to be diverse across the plexin family. Plexin-B1 and –B3 TM helices are regular and tend to associate, whereas plexin-A1, -A2, –A3, -A4, -C1 and –D1 contain sequence elements, such as poly-Glycine or aromatic residues that distort helix conformation and association. Plexin-B2 does not form stable dimers due to the presence of TM prolines. No experimental structural information on the TM region is available for these proteins, except for plexin-C1 dimeric and plexin-B1 – trimeric structures inferred from X-ray crystal structures of the intracellular regions. Plexin-B1 TM trimers utilize Ser and Thr sidechains for interhelical contacts. We also modeled the juxta-membrane (JM) region of plexin-C1 and plexin-B1 and show that it synergizes with the TM structures. The structure and dynamics of the JM region and TM-JM junction provide determinants for the distance and distribution of the intracellular domains, and for their binding partners relative to the membrane. The structures suggest experimental tests and will be useful for the interpretation of future studies.     

Towards a structural understanding of the smallest known oncoprotein: investigation of the bovine papillomavirus E5 protein using solution-state NMR

The homodimeric E5 protein from bovine papillomavirus activates the platelet-derived growth factor β receptor through transmembrane (TM) helix-helix interactions leading to uncontrolled cell growth. Detailed structural information for the E5 dimer is essential if we are to uncover its unique mechanism of action. In vivo mutagenesis has been used to identify residues in the TM domain critical for dimerization, and we previously reported that a truncated synthetic E5 peptide forms dimers via TM domain interactions. Here we extend this work with the first application of high-resolution solution-state NMR to the study of the E5 TM domain in SDS micelles. Using selectively 15N-labelled peptides, we first probe sample homogeneity revealing two predominate species, which we interpret to be monomer and dimer. The equilibrium between the two states is shown to be dependent on detergent concentration, revealed by intensity shifts between two sets of peaks in 15N-(1)H HSQC experiments, highlighting the importance of sample preparation when working with these types of proteins. This information is used to estimate a free energy of association (ΔGx°=-3.05 kcal mol(-1)) for the dimerization of E5 in SDS micelles. In addition, chemical shift changes have been observed that indicate a more pronounced change in chemical environment for those residues expected to be at the dimer interface in vivo versus those that are not. Thus we are able to demonstrate our in vitro dimer is comparable to that defined in vivo, validating the biological significance of our synthetic peptide and providing a solid foundation upon which to base further structural studies. Using detergent concentration to modulate oligomeric state and map interfacial residues by NMR could prove useful in the study of other homo-oligomeric transmembrane proteins.     

A photon-free approach to transmembrane protein structure determination

The structures of membrane proteins are generally solved using samples dissolved in micelles, bicelles, or occasionally phospholipid bilayers using X-ray diffraction or magnetic resonance. Because these are less than perfect mimics of true biological membranes, the structures are often confirmed by evaluating the effects of mutations on the properties of the protein in their native cellular environments. Low-resolution structures are also sometimes generated from the results of site-directed mutagenesis when other structural data are incomplete or not available. Here, we describe a rapid and automated approach to determine structures from data on site-directed mutants for the special case of homo-oligomeric helical bundles. The method uses as input an experimental profile of the effects of mutations on some property of the protein. This profile is then interpreted by assuming that positions that have large effects on structure/function when mutated project toward the center of the oligomeric bundle. Model bundles are generated, and correlation analysis is used to score which structures have inter-subunit C^β distances between adjoining monomers that best correlate with the experimental profile. These structures are then clustered and refined using energy-based minimization methods. For a set of 10 homo-oligomeric TM protein structures ranging from dimers to pentamers, we show that our method predicts structures to within 1-2 Å backbone RMSD relative to X-ray and NMR structures. This level of agreement approaches the precision of NMR structures solved in different membrane mimetics.     

Local energy landscape flattening: parallel hyperbolic Monte Carlo sampling of protein folding

Among the major difficulties in protein structure prediction is the roughness of the energy landscape that must be searched for the global energy minimum. To address this issue, we have developed a novel Monte Carlo algorithm called parallel hyperbolic sampling (PHS) that logarithmically flattens local high-energy barriers and, therefore, allows the simulation to tunnel more efficiently through energetically inaccessible regions to low-energy valleys. Here, we show the utility of this approach by applying it to the SICHO (SIde-CHain-Only) protein model. For the same CPU time, the parallel hyperbolic sampling method can identify much lower energy states and explore a larger region phase space than the commonly used replica sampling (RS) Monte Carlo method. By clustering the simulated structures obtained in the PHS implementation of the SICHO model, we can successfully predict, among a representative benchmark 65 proteins set, 50 cases in which one of the top 5 clusters have a root-mean-square deviation (RMSD) from the native structure below 6.5 A. Compared with our previous calculations that used RS as the conformational search procedure, the number of successful predictions increased by four and the CPU cost is reduced. By comparing the structure clusters produced by both PHS and RS, we find a strong correlation between the quality of predicted structures and the minimum relative RMSD (mrRMSD) of structures clusters identified by using different search engines. This mrRMSD correlation may be useful in blind prediction as an indicator of the likelihood of successful folds.     

Hydrophobic side-chain size is a determinant of the three-dimensional structure of the p53 oligomerization domain.

The p53 tumor suppressor oligomerization domain, a dimer of two primary dimers, is an independently folding domain whose subunits consist of a beta-strand, a tight turn and an alpha-helix. To evaluate the effect of hydrophobic side-chains on three-dimensional structure, we substituted residues Phe341 and Leu344 in the alpha-helix with other hydrophobic amino acids. Substitutions that resulted in residue 341 having a smaller side-chain than residue 344 switched the stoichiometry of the domain from tetrameric to dimeric. The three-dimensional structure of one such dimer was determined by multidimensional NMR spectroscopy. When compared with the primary dimer of the wild-type p53 oligomerization domain, the mutant dimer showed a switch in alpha-helical packing from anti-parallel to parallel and rotation of the alpha-helices relative to the beta-strands. Hydrophobic side-chain size is therefore an important determinant of a protein fold.     

The dimeric transmembrane domain of prolyl dipeptidase DPP-IV contributes to its quaternary structure and enzymatic activities

Dipeptidyl peptidase IV (DPP‐IV) is a drug target in the treatment of human type II diabetes. It is a type II membrane protein with a single transmembrane domain (TMD) anchoring the extracellular catalytic domain to the membrane. DPP‐IV is active as a dimer, with two dimer interacting surfaces located extracellularly. In this study, we demonstrate that the TM of DPP‐IV promotes DPP‐IV dimerization and rescues monomeric DPP‐IV mutants into partial dimers, which is specific and irreplaceable by TMs of other type II membrane proteins. By bioluminescence resonance energy transfer (BRET) and peptide electrophoresis, we found that the TM domain of DPP‐IV is dimerized in mammalian cells and in vitro. The TM dimer interaction is very stable, based on our results with TM site‐directed mutagenesis. None of the mutations, including the introduction of two prolines, resulted in their complete disruption to monomers. However, these TM proline mutations result in a significant reduction of DPP‐IV enzymatic activity, comparable to what is found with mutations near the active site. A systematic analysis of TM structures deposited in the Protein Data Bank showed that prolines in the TM generally produce much bigger kinking angles than occur in nonproline‐containing TMs. Thus, the proline‐dependent reduction in enzyme activity may result from propagated conformational changes from the TM to the extracellular active site. Our results demonstrate that TM dimerization and conformation contribute significantly to the structure and activity of DPP‐IV. Optimal enzymatic activity of DPP‐IV requires an optimal interaction of all three dimer interfaces, including its TM.     

Molecular dynamics simulation approach for the prediction of transmembrane helix–helix heterodimers assembly

Computational methods are useful to identify favorable structures of transmembrane (TM) helix oligomers when experimental data are not available or when they cannot help to interpret helix-helix association. We report here a global search method using molecular dynamics (MD) simulations to predict the structures of transmembrane homo and heterodimers. The present approach is based only on sequence information without any experimental data and is first applied to glycophorin A to validate the protocol and to the HER2-HER3 heterodimer receptor. The method successfully reproduces the experimental structures of the TM domain of glycophorin A (GpA(TM)) with a root mean square deviation of 1.5 A. The search protocol identifies three energetically stable models of the TM domain of HER2-HER3 receptor with favorable helix-helix arrangement, including right-handed and left-handed coiled-coils. The predicted TM structures exhibit the GxxxG-like motif at the dimer interface which is presumed to drive receptor oligomerization. We demonstrate that native structures of TM domain can be predicted without quantitative experimental data. This search protocol could help to predict structures of the TM domain of HER heterodimer family.     

Retention of native-like oligomerization states in transmembrane segment peptides: application to the Escherichia coli aspartate receptor

Biophysical study of the transmembrane (TM) domains of integral membrane proteins has traditionally been impeded by their hydrophobic nature. As a result, an understanding of the details of protein-protein interactions within membranes is often lacking. We have demonstrated previously that model TM segments with flanking cationic residues spontaneously fold into alpha-helices upon insertion into membrane-mimetic environments. Here, we extend these studies to investigate whether such constructs consisting of TM helices from biological systems retain their native secondary structures and oligomeric states. Single-spanning TM domains from the epidermal growth factor receptor (EGFR), glycophorin A (GPA), and the influenza A virus M2 ion channel (M2) were designed and synthesized with three to four lysine residues at both N- and C-termini. Each construct was shown to adopt an alpha-helical conformation upon insertion into sodium dodecyl sulfate micelles. Furthermore, micelle-inserted TM segments associated on SDS-PAGE gels according to their respective native-like oligomeric states: EGFR was monomeric, GPA was dimeric, and M2 was tetrameric. This approach was then used to investigate whether one or both of the TM segments (Tar-1 and Tar-2) from the Escherichia coli aspartate receptor were responsible for its homodimeric nature. Our results showed that Tar-1 formed SDS-resistant homodimers, while Tar-2 was monomeric. Furthermore, no heterooligomerization between Tar-1 and Tar-2 was detected, implicating the Tar-1 helix as the oligomeric determinant for the Tar protein. The overall results indicate that this approach can be used to elucidate the details of TM domain folding for both single-spanning and multispanning membrane proteins.     

TOUCHSTONE II: a new approach to ab initio protein structure prediction

We have developed a new combined approach for ab initio protein structure prediction. The protein conformation is described as a lattice chain connecting C(alpha) atoms, with attached C(beta) atoms and side-chain centers of mass. The model force field includes various short-range and long-range knowledge-based potentials derived from a statistical analysis of the regularities of protein structures. The combination of these energy terms is optimized through the maximization of correlation for 30 x 60,000 decoys between the root mean square deviation (RMSD) to native and energies, as well as the energy gap between native and the decoy ensemble. To accelerate the conformational search, a newly developed parallel hyperbolic sampling algorithm with a composite movement set is used in the Monte Carlo simulation processes. We exploit this strategy to successfully fold 41/100 small proteins (36 approximately 120 residues) with predicted structures having a RMSD from native below 6.5 A in the top five cluster centroids. To fold larger-size proteins as well as to improve the folding yield of small proteins, we incorporate into the basic force field side-chain contact predictions from our threading program PROSPECTOR where homologous proteins were excluded from the data base. With these threading-based restraints, the program can fold 83/125 test proteins (36 approximately 174 residues) with structures having a RMSD to native below 6.5 A in the top five cluster centroids. This shows the significant improvement of folding by using predicted tertiary restraints, especially when the accuracy of side-chain contact prediction is >20%. For native fold selection, we introduce quantities dependent on the cluster density and the combination of energy and free energy, which show a higher discriminative power to select the native structure than the previously used cluster energy or cluster size, and which can be used in native structure identification in blind simulations. These procedures are readily automated and are being implemented on a genomic scale.     

Specificity of helix packing in transmembrane dimer of the cell death factor BNIP3: a molecular modeling study

BNIP3 is a mitochondrial 19-kDa proapoptotic protein, a member of the Bcl-2 family. It has a single COOH-terminal transmembrane (TM) alpha-helical domain, which is required for membrane targeting, proapoptotic activity, hetero- and homo-dimerization in membrane. The role and the molecular details of association of TM helices of BNIP3 are yet to be established. Here, we present a molecular modeling study of helix interactions in its membrane domain. The approach combines Monte Carlo conformational search in an implicit hydrophobic slab followed by molecular dynamics simulations in a hydrated full-atom lipid bilayer. The former technique was used for exhaustive sampling of the peptides' conformational space and for generation of putative "native-like" structures of the dimer. The latter ones were taken as realistic starting points to assess stability and dynamic behavior of the complex in explicit lipid-water surrounding. As a result, several groups of tightly packed right-handed structures of the dimer were proposed. They have almost similar helix-helix interface, which includes the motif A(176)xxxG(180)xxxG(184) and agrees well with previous mutagenesis data and preliminary NMR analysis. Molecular dynamics simulations of these structures reveal perfect adaptation of most of them to heterogeneous membrane environment. A remarkable feature of the predicted dimeric structures is the occurrence of a cluster of H-bonded histidine 173 and serines 168 and 172 on the helix interface, near the N-terminus. Because of specific polar interactions between the monomers, this part of the dimer has no such dense packing as the C-terminal one, thus allowing penetration of water from the extramembrane side into the membrane interior. We propose that the ionization state of His(173) can mediate structural and dynamic properties of the dimer. This, in turn, may be related to pH-dependent proapoptotic activity of BNIP3, which is triggering on by acidosis appearing under hypoxic conditions.     

A de novo designed protein protein interface

As an approach to both explore the physical/chemical parameters that drive molecular self-assembly and to generate novel protein oligomers, we have developed a procedure to generate protein dimers from monomeric proteins using computational protein docking and amino acid sequence design. A fast Fourier transform-based docking algorithm was used to generate a model for a dimeric version of the 56-amino-acid beta1 domain of streptococcal protein G. Computational amino acid sequence design of 24 residues at the dimer interface resulted in a heterodimer comprised of 12-fold and eightfold variants of the wild-type protein. The designed proteins were expressed, purified, and characterized using analytical ultracentrifugation and heteronuclear NMR techniques. Although the measured dissociation constant was modest ( approximately 300 microM), 2D-[(1)H,(15)N]-HSQC NMR spectra of one of the designed proteins in the absence and presence of its binding partner showed clear evidence of specific dimer formation.     

Assembly of transmembrane helices of simple polytopic membrane proteins from sequence conservation patterns

The transmembrane (TM) domains of most membrane proteins consist of helix bundles. The seemingly simple task of TM helix bundle assembly has turned out to be extremely difficult. This is true even for simple TM helix bundle proteins, i.e., those that have the simple form of compact TM helix bundles. Herein, we present a computational method that is capable of generating native-like structural models for simple TM helix bundle proteins having modest numbers of TM helices based on sequence conservation patterns. Thus, the only requirement for our method is the presence of more than 30 homologous sequences for an accurate extraction of sequence conservation patterns. The prediction method first computes a number of representative well-packed conformations for each pair of contacting TM helices, and then a library of tertiary folds is generated by overlaying overlapping TM helices of the representative conformations. This library is scored using sequence conservation patterns, and a subsequent clustering analysis yields five final models. Assuming that neighboring TM helices in the sequence contact each other (but not that TM helices A and G contact each other), the method produced structural models of Calpha atom root-mean-square deviation (CA RMSD) of 3-5 A from corresponding crystal structures for bacteriorhodopsin, halorhodopsin, sensory rhodopsin II, and rhodopsin. In blind predictions, this type of contact knowledge is not available. Mimicking this, predictions were made for the rotor of the V-type Na(+)-adenosine triphosphatase without such knowledge. The CA RMSD between the best model and its crystal structure is only 3.4 A, and its contact accuracy reaches 55%. Furthermore, the model correctly identifies the binding pocket for sodium ion. These results demonstrate that the method can be readily applied to ab initio structure prediction of simple TM helix bundle proteins having modest numbers of TM helices.     

Tertiary structure predictions on a comprehensive benchmark of medium to large size proteins

We evaluate tertiary structure predictions on medium to large size proteins by TASSER, a new algorithm that assembles protein structures through rearranging the rigid fragments from threading templates guided by a reduced Calpha and side-chain based potential consistent with threading based tertiary restraints. Predictions were generated for 745 proteins 201-300 residues in length that cover the Protein Data Bank (PDB) at the level of 35% sequence identity. With homologous proteins excluded, in 365 cases, the templates identified by our threading program, PROSPECTOR_3, have a root-mean-square deviation (RMSD) to native < 6.5 angstroms, with >70% alignment coverage. After TASSER assembly, in 408 cases the best of the top five full-length models has a RMSD < 6.5 angstroms. Among the 745 targets are 18 membrane proteins, with one-third having a predicted RMSD < 5.5 A. For all representative proteins less than or equal to 300 residues that have corresponding multiple NMR structures in the Protein Data Bank, approximately 20% of the models generated by TASSER are closer to the NMR structure centroid than the farthest individual NMR model. These results suggest that reasonable structure predictions for nonhomologous large size proteins can be automatically generated on a proteomic scale, and the application of this approach to structural as well as functional genomics represent promising applications of TASSER.     

Energetics of the native and non-native states of the glycophorin transmembrane helix dimer

Using an implicit membrane model (IMM1), we examine whether the structure of the transmembrane domain of Glycophorin A (GpA) could be predicted based on energetic considerations alone. The energetics of native GpA shows that van der Waals interactions make the largest contribution to stability. Although specific electrostatic interactions are stabilizing, the overall electrostatic contribution is close to zero. The GXXXG motif contributes significantly to stability, but residues outside this motif contribute almost twice as much. To generate non-native states a global conformational search was done on two segments of GpA: an 18-residue peptide (GpA74-91) that is embedded in the membrane and a 29-residue peptide (GpA70-98) that has additional polar residues flanking the transmembrane region. Simulated annealing was done on a large number of conformations generated from parallel, antiparallel, left- and right-handed starting structures by rotating each helix at 20 degrees intervals around its helical axis. Several crossing points along the helix dimer were considered. For 18-residue parallel topology, an ensemble of native-like structures was found at the lowest effective energy region; the effective energy is lowest for a right-handed structure with an RMSD of 1.0 A from the solid-state NMR structure with correct orientation of the helices. For the 29-residue peptide, the effective energies of several left-handed structures were lower than that of the native, right-handed structure. This could be due to deficiencies in modeling the interactions between charged sidechains and/or omission of the sidechain entropy contribution to the free energy. For 18-residue antiparallel topology, both IMM1 and a Generalized Born model give effective energies that are lower than that of the native structure. In contrast, the Poisson-Boltzmann solvation model gives lower effective energy for the parallel topology, largely because the electrostatic solvation energy is more favorable for the parallel structure. IMM1 seems to underestimate the solvation free energy advantage when the CO and NH dipoles just outside the membrane are parallel. This highlights the importance of electrostatic interactions even when these are not obvious by looking at the structures.     

Antiparallel dimers of the small multidrug resistance protein EmrE are more stable than parallel dimers

The bacterial multidrug transporter EmrE is a dual-topology membrane protein and as such is able to insert into the membrane in two opposite orientations. The functional form of EmrE is a homodimer; however, the relative orientation of the subunits in the dimer is under debate. Using EmrE variants with fixed, opposite orientations in the membrane, we now show that, although the proteins are able to form parallel dimers, an antiparallel organization of the subunits in the dimer is preferred. Blue-native PAGE analyses of intact oligomers and disulfide cross-linking demonstrate that in membranes, the proteins form parallel dimers only if no oppositely orientated partner is present. Co-expression of oppositely orientated proteins almost exclusively yields antiparallel dimers. Finally, parallel dimers can be disrupted and converted into antiparallel dimers by heating of detergent-solubilized protein. Importantly, in vivo function is correlated clearly to the presence of antiparallel dimers. Our results suggest that an antiparallel arrangement of the subunits in the dimer is more stable than a parallel organization and likely corresponds to the functional form of the protein.     

Conformational pathway for the kissing complex-->extended dimer transition of the SL1 stem-loop from genomic HIV-1 RNA as monitored by targeted molecular dynamics techniques

HIV-1 retroviral genomic RNA dimerization is initiated by loop-loop interactions between the SL1 stem-loops of two identical RNA molecules. The SL1-SL1 unstable resulting kissing complex (KC) then refolds irreversibly into a more stable complex called extended dimer (ED). Although the structures of both types of complex have been determined, very little is known about the conformational pathway corresponding to the transition, owing to the difficulty of observing experimentally intermediate conformations. In this study, we applied targeted molecular dynamics simulation techniques (TMD) to the phosphorus atoms for monitoring this pathway for the backbone, and a two-step strategy was adopted. In a first step, called TMD(-1), the dimer structure was constrained to progressively move away from KC without indicating the direction, until the RMSD from KC reaches 36A. A total of 20 TMD(-1) simulations were performed under different initial conditions and different simulation parameters. For RMSD ranging between 0 and 22A, the whole set of TMD(-1) simulations follows a similar pathway, then divergences are observed. None of the simulations leads to the ED structure. At RMSD=22A, the dimers look like two parallel Us, still linked by the initial loop-loop interaction, but the strands of the stems (the arms of the Us) are positioned in such a manner that they can form intramolecular as well as intermolecular Watson-Crick base-pairs. This family of structure is called UU. In a second step (TMD simulations), 18 structures were picked up along the pathways generated with TMD(-1) and were constrained to move toward ED by decreasing progressively their RMSD from ED. We found that only structures from the UU family are able to easily reach ED-like conformations of the backbones without exhibiting a large constraint energy.     

Proline-induced distortions of transmembrane helices

Proline residues in the transmembrane (TM) alpha-helices of integral membrane proteins have long been suspected to play a key role for helix packing and signal transduction by inducing regions of helix distortion and/or dynamic flexibility (hinges). In this study we try to characterise the effect of proline on the geometric properties of TM alpha-helices. We have examined 199 transmembrane alpha-helices from polytopic membrane proteins of known structure. After examining the location of proline residues within the amino acid sequences of TM helices, we estimated the helix axes either side of a hinge and hence identified a hinge residue. This enabled us to calculate helix kink and swivel angles. The results of this analysis show that proline residues occur with a significant concentration in the centre of sequences of TM alpha-helices. In this location, they may induce formation of molecular hinges, located on average about four residues N-terminal to the proline residue. A superposition of proline-containing TM helices structures shows that the distortion induced is anisotropic and favours certain relative orientations (defined by helix kink and swivel angles) of the two helix segments.     

TASSER-based refinement of NMR structures

The TASSER structure prediction algorithm is employed to investigate whether NMR structures can be moved closer to their corresponding X-ray counterparts by automatic refinement procedures. The benchmark protein dataset includes 61 nonhomologous proteins whose structures have been determined by both NMR and X-ray experiments. Interestingly, by starting from NMR structures, the majority (79%) of TASSER refined models show a structural shift toward their X-ray structures. On average, the TASSER refined models have a root-mean-square-deviation (RMSD) from the X-ray structure of 1.785 A (1.556 A) over the entire chain (aligned region), while the average RMSD between NMR and X-ray structures (RMSD(NMR_X-ray)) is 2.080 A (1.731 A). For all proteins having a RMSD(NMR_X-ray) >2 A, the TASSER refined structures show consistent improvement. However, for the 34 proteins with a RMSD(NMR_X-ray) <2 A, there are only 21 cases (60%) where the TASSER model is closer to the X-ray structure than NMR, which may be due to the inherent resolution of TASSER. We also compare the TASSER models with 12 NMR models in the RECOORD database that have been recalculated recently by Nederveen et al. from original NMR restraints using the newest molecular dynamics tools. In 8 of 12 cases, TASSER models show a smaller RMSD to X-ray structures; in 3 of 12 cases, where RMSD(NMR_X-ray) <1 A, RECOORD does better than TASSER. These results suggest that TASSER can be a useful tool to improve the quality of NMR structures.