Global folding of proteins using a limited number of distance constraints

A Monte Carlo method is presented which can obtain the correcttertiary fold of a protein given the secondary structure andas few as three interactions between each secondary structureunit. This method was used to fold hemerythrin, Qavodoxin, bovinepancreatic trypsin inhibitor and a variable light domain froman immunoglobulin using the known secondary structures of theseproteins. Each of the proteins was successfully folded to obtaina structure resembling the initial X-ray structure. Reasonablesuccess was also achieved when using a secondary structure predictionalgorithm to assign secondary structure. The r.m.s. deviationsbetween the folded proteins and the crystal structures are inthe order of 3–5 A for the backbone coordinates. Evaluationof the r.m.s. deviations between members of the globin familyindicates that two equivalent overall folds may have r.m.s.deviations of this or even larger magnitude. The limiting numberof constraints necesssary to achieve the correct fold is discussed.     

Recognizing misfolded and distorted protein structures by the assumption-based similarity score

A new similarity score (-score) is proposed which is able tofind the correct protein structure among the very close alternativesand to distinguish between correct and deliberately misfoldedstructures. This score is based on the general principle `similarlikes similar', and it favors hydrophobic and hydrophilic contacts,and disfavors hydrophobic-to-hydrophilic contacts in proteins.The values of -scores calculated for the high-resolution proteinstructures from the representative set are compared with thoseof alternatives: (i) very close alternatives which are onlyslightly distorted by conformational energy minimization invacuo; (ii) alternatives with subsequently growing distortions,generated by molecular dynamics simulations in vacuo; (iii)structures derived by molecular dynamics simulation in solventat 300 K; (iv) deliberately misfolded protein models. In nearlyall tested cases the similarity score can successfully distinguishbetween experimental structure and its alternatives, even ifthe root mean square displacement of all heavy atoms is lessthan 1 Å. The confidence interval of the similarity scorewas estimated using the high-resolution X-ray structures ofdomain pairs related by non-crystallographic symmetry. The similarityscore can be used for the evaluation of the general qualityof the protein models, choosing the correct structures amongthe very close alternatives, characterization of models simulatingfolding/unfolding, etc.     

Local polarity analysis: a sensitive method that discriminates between native proteins and incorrectly folded models

The evaluation of calculated protein structures is an importantstep in the protein design cycle. Known criteria for this assessmentof proteins are the polar and apolar, accessible and buriedsurface area, electrostatic interactions and other interactionsbetween the protein atoms (e.g. HO, S-S),atomic packing, analysisof amino acid environment and surface charge distribution. Weshow that a powerful test of accuracy of protein structure canbe derived by analysing the water contact of atoms and additionallytaking into account their polarity. On the basis of estimatedreference values of the polar fraction of typical globular proteinswith known structure (mean, SD and distribution), the evaluationof misfolded structures can be improved significantly. The referencevalues are derived by moving windows of different length (3–99amino acid residues) over the amino acid sequence. Model proteins,which are included in the Brookhaven protein structure databank,deliberately misfolded proteins, hypothetical proteins and predictedprotein structures are diagnosed as at least partially incorrectlyfolded. The local fault, mostly observed, is that polar groupsare buried too frequently in the interior of the protein. Thedatabase-derived quantities are useful in screening the designedproteins prior to experimentation and may also be useful inthe assessment of errors in the experimentally determined proteinstructures.     

Role of prion protein aggregation in neurotoxicity

In several neurodegenerative diseases, such as Parkinson, Alzheimer's, Huntington, and prion diseases, the deposition of aggregated misfolded proteins is believed to be responsible for the neurotoxicity that characterizes these diseases. Prion protein (PrP), the protein responsible of prion diseases, has been deeply studied for the peculiar feature of its misfolded oligomers that are able to propagate within affected brains, inducing the conversion of the natively folded PrP into the pathological conformation. In this review, we summarize the available experimental evidence concerning the relationship between aggregation status of misfolded PrP and neuronal death in the course of prion diseases. In particular, we describe the main findings resulting from the use of different synthetic (mainly PrP106-126) and recombinant PrP-derived peptides, as far as mechanisms of aggregation and amyloid formation, and how these different spatial conformations can affect neuronal death. In particular, most data support the involvement of non-fibrillar oligomers rather than actual amyloid fibers as the determinant of neuronal death.     

The Role of SBI2/ALG12/EBS4 in the Regulation of Endoplasmic Reticulum-Associated Degradation (ERAD) Studied by a Null Allele

Redundancy and lethality is a long-standing problem in genetics but generating minimal and lethal phenotypes in the knockouts of the same gene by different approaches drives this problem to a new high. In Asn (N)-linked glycosylation, a complex and ubiquitous cotranslational and post-translational protein modification required for the transfer of correctly folded proteins and endoplasmic reticulum-associated degradation (ERAD) of misfolded proteins, ALG12 (EBS4) is an α 1, 6-mannosyltransferase catalyzing a mannose into Glc₃Man₉GlcNAc₂. In Arabidopsis, T-DNA knockout alg12-T is lethal while likely ebs4 null mutants isolated by forward genetics are most healthy as weak alleles, perplexing researchers and demanding further investigations. Here, we isolated a true null allele, sbi2, with the W258Stop mutation in ALG12/EBS4. sbi2 restored the sensitivity of brassinosteroid receptor mutants bri1-5, bri1-9, and bri1-235 with ER-trapped BRI1 to brassinosteroids. Furthermore, sbi2 maturated earlier than the wild-type. Moreover, concomitant with impaired and misfolded proteins accumulated in the ER, sbi2 had higher sensitivity to tunicamycin and salt than the wild-type. Our findings thus clarify the role of SBI2/ALG12/EBS4 in the regulation of the ERAD of misfolded glycoproteins, and plant growth and stress response. Further, our study advocates the necessity and importance of using multiple approaches to validate genetics study.     

Network pattern of residue packing in helical membrane proteins and its application in membrane protein structure prediction

De novo protein structure prediction plays an important role in studies of helical membrane proteins as well as structure-based drug design efforts. Developing an accurate scoring function for protein structure discrimination and validation remains a current challenge. Network approaches based on overall network patterns of residue packing have proven useful in soluble protein structure discrimination. It is thus of interest to apply similar approaches to the studies of residue packing in membrane proteins. In this work, we first carried out such analysis on a set of diverse, non-redundant and high-resolution membrane protein structures. Next, we applied the same approach to three test sets. The first set includes nine structures of membrane proteins with the resolution worse than 2.5 A; the other two sets include a total of 101 G-protein coupled receptor models, constructed using either de novo or homology modeling techniques. Results of analyses indicate the two criteria derived from studying high-resolution membrane protein structures are good indicators of a high-quality native fold and the approach is very effective for discriminating native membrane protein folds from less-native ones. These findings should be of help for the investigation of the fundamental problem of membrane protein structure prediction.     

A relational database of protein structures designed for flexible enquiries about conformation

A relational database of protein structure has been developedto enable rapid and flexible enquiries about the occurrenceof many aspects of protein architecture. The coordinates of294 proteins from the Brookhaven Data Bank have been processedby standard computer programs to generate many additional termsthat quantify aspects of protein structure. These terms includesolvent accessibility, main-chain and side-chain dihedral angles,and secondary structure. In a relational database, the informationis stored in tables with columns holding the different termsand rows holding the different entries for the terms. The differentrelational base tables store the information about the proteincoordinate set, the different chains in the protein, the aminoacid residues and ligands, the atomic coordinates, the saltbridges, the hydrogen bonds, the disulphide bridges and theclose tertiary contacts. The database was established underORACLE management system. Enquiries are constructed in ORACLEusing SQL (structured query language) which is simple to useand alleviates the need for extensive computer programs. A singletable can be searched for entries that meet various criteria,e.g. all protein solved to better than a given resolution. Thepower of the database occurs when several tables, or the entriesin a single table, are cross-correlated. For example the dihedralangles of proline in the fourth position in an -helix in highresolution structures can be rapidly obtained. The structuraldatabase provides a powerful tool to obtain empirical rulesabout protein conformation. This database of protein structuresis part of a joint project between Birkbeck College and LeedsUniversity to establish an integrated data resource of proteinsequences and structures (ISIS) that encodes the complex patternsof residues and coordinates that define protein conformation.The entire data resource (ISIS) will provide a system to guideall areas of protein modelling including structure prediction,site-directed mutagenesis and de novo protein design. The availabilityof ISIS is described in the paper.     

A de Novo‐Designed Monomeric,Compact Three‐Helix‐Bundle Protein on a Carbohydrate Template

De novo design and chemical synthesis of proteins and of other artificial structures that mimic them is a central strategy for understanding protein folding and for accessing proteins with new functions. We have previously described carbohydrates that act as templates for the assembly of artificial proteins, so‐called carboproteins. The hypothesis is that the template preorganizes the secondary structure elements and directs the formation of a tertiary structure, thus achieving structural economy in the combination of peptide, linker, and template. We speculate that the structural information from the template could facilitate protein folding. Here we report the design and synthesis of three‐helix‐bundle carboproteins on deoxyhexopyranosides. The carboproteins were analyzed by CD, analytical ultracentrifugation (AUC), small‐angle X‐ray scattering (SAXS), and NMR spectroscopy, and this revealed the formation of the first compact and folded monomeric carboprotein, distinctly different from a molten globule. En route to this carboprotein we observed a clear effect originating from the template on protein folding.     

Predicting surface exposure of amino acids from protein sequence

The amino acid residues on a protein surface play a key rolein interaction with other molecules, determine many physicalproperties, and constrain the structure of the folded protein.A database of monomeric protein crystal structures was usedto teach computer-simulated neural networks rules for predictingsurface exposure from local sequence. These trained networksare able to correctly predict surface exposure for 72% of residuesin a testing set using a binary model (buried/exposed) and for54% of residues using a ternary model (buried/intermediate/exposed).In the ternary model, only 11% of the exposed residues are predictedas buried and only 5% of the buried residues are predicted asexposed. Also, since the networks are able to predict exposurewith a quantitative confidence estimate, it is possible to assignexposure for over half of the residues in a binary model with>80% accuracy. Even more accurate predictions are obtainedby making a consensus prediction of exposure for a homologousfamily. The effect of the local environment of an amino acidon its accessibility, though smaller than expected, is significantand accounts for the higher success rate of prediction thanobtained with previously used criteria. In the absence of athree-dimensional structure, the ability to predict surfaceaccessibility of amino acids directly from the sequence is avaluable tool in choosing sites of chemical modification orspecific mutations and in studies of molecular interaction.     

Alignment of protein sequences using secondary structure: a modified dynamic programming method

A method for comparison of protein sequences based on theirprimary and secondary structure is described. Protein sequencesare annotated with predicted secondary structures (using a modifiedChou and Fasman method). Two lettered code sequences are generated(Xx, where X is the amino acid and x is its annotated secondarystructure). Sequences are compared with a dynamic programmingmethod (STRALIGN) that includes a similarity matrix for boththe amino acids and secondary structures. The similarity valuefor each paired two-lettered code is a linear combination ofsimilarity values for the paired amino acids and their annotatedsecondary structures. The method has been applied to eight globinproteins (28 pairs) for which the X-ray structure is known.For protein pairs with high primary sequence similarity (>45%),STRALIGN alignment is identical to that obtained by a dynamicprogramming method using only primary sequence information.However, alignment of protein pairs with lower primary sequencesimilarity improves significantly with the addition of secondarystructure annotation. Alignment of the pair with the least primarysequence similarity of 16% was improved from 0 to 37% ‘correct’alignment using this method. In addition, STRALIGN was successfullyapplied to seven pairs of distantly related cytochrome c proteins,and three pairs of distantly related picornavirus proteins.     

Symmetry‐Directed Self‐Assembly of a Tetrahedral Protein Cage Mediated by de Novo‐Designed Coiled Coils

The organization of proteins into new hierarchical forms is an important challenge in synthetic biology. However, engineering new interactions between protein subunits is technically challenging and typically requires extensive redesign of protein–protein interfaces. We have developed a conceptually simple approach, based on symmetry principles, that uses short coiled‐coil domains to assemble proteins into higher‐order structures. Here, we demonstrate the assembly of a trimeric enzyme into a well‐defined tetrahedral cage. This was achieved by genetically fusing a trimeric coiled‐coil domain to its C terminus through a flexible polyglycine linker sequence. The linker length and coiled‐coil strength were the only parameters that needed to be optimized to obtain a high yield of correctly assembled protein cages.     

Neural network system for the evaluation of side-chain packing in protein structures

An artificial neural network system is used for pattern recognitionin protein side-chain-side-chain contact maps. A back-propagationnetwork was trained on a set of patterns which are popular inside-chain contact maps of protein structures. Several neuralnetwork architectures and different training parameters weretested to decide on the best combination for the neural network.The resulting network can distinguish between original (fromprotein structures) and randomized patterns with an accuracyof 84.5% and a Matthews' coefficient of 0.72 for the testingset. Applications of this system for protein structure evaluationand refinement are also proposed. Examples include structuresobtained after the application of molecular dynamics to crystalstructures, structures obtained from X-ray crystallography atvarious stages of refinement, structures obtained from a denovo folding algorithm and deliberately misfolded structures.     

Normal Mode Dynamics Comparison of Proteins

Sequence and structure comparisons are fundamental techniques that enable exploration of the sequence and structural spaces of proteins. Homology detection, function prediction, and protein classification rely on these techniques. However, protein structures are dynamic, rather than static, and such protein dynamics play a key role in a wide range of biological activities. Therefore, protein dynamics comparison algorithms may shed light on the relationship between proteins′ dynamics and function. Here, we review different strategies for comparing dynamics of proteins. Special emphasis is given to newly developed algorithms that compare dynamics of proteins with no apparent sequence or structural similarity and to the qualitative differences between these algorithms.     

Polarity as a criterion in protein design

Hypothetical proteins can be tested computationally by determiningwhether or not the designed sequence-structure pair has thecharacteristics of a typical globular protein. We have developedsuch a test by deriving quantities with approximately constantvalue for all globular proteins, based on empirical analysisof the exposed and buried surfaces of 128 structurally knownproteins. The characteristic quantities that best appear tosegregate badly designed or deliberately misfolded proteinsfrom their properly folded natural relatives are the polar fractionof side chains on the protein surface and, independently, inthe protein interior. Three of the seven hypothetical structurestested here can be rejected as having too many polar side-chaingroups in the interior or too few on the protein surface. Inaddition, a recently designed nutritional protein is identifiedas being very much unlike globular proteins. These database-derivedcharacteristic quantities are useful in screening designed proteinsprior to experiment and may be useful in screening experimentallydetermined (X-ray, NMR) protein structures for possible errors.     

Nanoscale Structural Organization of Insulin Fibril Polymorphs Revealed by Atomic Force Microscopy–Infrared Spectroscopy (AFM-IR)

Spontaneous aggregation of misfolded proteins typically results in the formation of morphologically and structurally different amyloid fibrils, protein aggregates that are strongly associated with various neurodegenerative disorders. Elucidation of the structural organization of amyloid aggregates is crucial to understanding their role in the onset and progression of these diseases. Using atomic force microscopy–infrared spectroscopy (AFM-IR), we investigated the structural organization of insulin fibrils. We found that insulin aggregation results in the formation of two structurally different fibril polymorphs. One polymorph has a β-sheet core surrounded by primarily unordered protein secondary structure. This polymorph has β-sheet-rich surface, whereas the surface of the other fibril polymorph is primarily composed of unordered protein. Using AFM-IR, we also revealed the structural organization of the insulin oligomers. Finally, we discovered a new pathway for amyloid fibril formation that is based on a fusion of several oligomers into a single fibril structure.     

Prediction of protein secondary structure by mining structural fragment database

A new method for predicting protein secondary structure from amino acid sequence has been developed. The method is based on multiple sequence alignment of the query sequence with all other sequences with known structure from the protein data bank (PDB) by using BLAST. The fragments of the alignments belonging to proteins from the PBD are then used for further analysis. We have studied various schemes of assigning weights for matching segments and calculated normalized scores to predict one of the three secondary structures: α-helix, β-sheet, or coil. We applied several artificial intelligence techniques: decision trees (DT), neural networks (NN) and support vector machines (SVM) to improve the accuracy of predictions and found that SVM gave the best performance. Preliminary data show that combining the fragment mining approach with GOR V (Kloczkowski et al, Proteins 49 (2002) 154-166) for regions of low sequence similarity improves the prediction accuracy.     

Limitations of yeast surface display in engineering proteins of high thermostability

Engineering proteins that can fold to unique structures remains a challenge. Protein stability has previously been engineered via the observed correlation between thermal stability and eukaryotic secretion level. To explore the limits of an expression-based approach, variants of the highly thermostable three-helix bundle protein alpha3D were studied using yeast surface display. A library of alpha3D mutants was created to explore the possible correlation of protein stability and fold with expression level. Five efficiently expressed mutants were then purified and further studied biochemically. Despite their differences in stability, most mutants expressed at levels comparable with that of wild-type alpha3D. Two other related sequences (alpha3A and alpha3B) that form collapsed, stable molten globules but lack a uniquely folded structure were similarly expressed at high levels by yeast display. Together these observations suggest that the quality control system in yeast is unable to discriminate between well-folded proteins of high stability and molten globules. The present study, therefore, suggests that an optimization of the surface display efficiency on yeast may yield proteins that are thermally and chemically stable yet are poorly folded.     

Beta-hairpin peptidomimetics: design, structures and biological activities

The folded 3D structures of peptides and proteins provide excellent starting points for the design of synthetic molecules that mimic key epitopes (or surface patches) involved in protein-protein and protein-nucleic acid interactions. Protein epitope mimetics (PEMs) may recapitulate not only the structural and conformational properties of the target epitope but also their biological activities. By transferring the epitope from a recombinant to a synthetic scaffold that can be produced by parallel combinatorial methods, it is possible to optimize properties through iterative cycles of library synthesis and screening, and even to evolve new biological activities. One very interesting scaffold is found in beta-hairpin motifs, which are used by many proteins to mediate molecular recognition events. This motif is readily amenable to PEM design, for example, by transplanting hairpin loop sequences from folded proteins onto hairpin-stabilizing templates, such as the dipeptide D-Pro-L-Pro. In addition, beta-hairpin peptidomimetics can also be exploited to mimic other types of epitopes, such as those based on alpha-helical secondary structures. The size and shape of beta-hairpin PEMs appear well suited for the design of inhibitors of both protein-protein and protein-nucleic acid interactions, endeavors that have so far proven difficult using small "drug-like" molecules. In recent work, it was shown that beta-hairpin PEMs can be designed that mimic the canonical conformations of antibody hypervariable loops, suggesting that novel small-molecule antibody mimics may be feasible. Using naturally occurring peptides as starting points, beta-hairpin mimetics have been discovered that possess antimicrobial activity, while others are potent inhibitors of the chemokine receptor CXCR4. Beta-hairpin PEMs have also been designed and optimized that mimic an alpha-helical epitope in p53 and so block its interaction with HDM2. A crystal structure of one HDM2-mimetic complex revealed how the surface of the protein had adapted to the shape of the hairpin, thereby enhancing inhibitor affinity. Small folded RNA motifs also make interesting targets for inhibitor design. For example, beta-hairpin mimetics have been designed and optimized that bind with high affinity and good selectivity to the TAR and RRE RNA motifs from HIV-1. Solution structures of the mimetics both free and bound to the RNA target provided some surprises, as well as an improved understanding of the mechanisms of binding. These mimetics represent still a relatively new family of RNA-binding molecules, but clearly one with potential for development into novel antiviral agents.     

The role of heat-shock and chaperone proteins in protein folding: possible molecular mechanisms

Recently some heat-shock proteins have been linked to functionsof ‘chaperoning’ protein folding in vivo. Here currentexperimental evidence is reviewed and possible requirementsfor such an activity are discussed. It is proposed that onemode of chaperone action is to actively unfold misfolded orbadly aggregated proteins to a conformation from whkh they couldrefold spontaneously; that improperly folded proteins are recognizedby excessive stretches of solvent-exposed backbone, rather thanby exposed hydrophobic patches; and that the molecular mechanismfor unfolding is either repeated binding and dissociation (‘plucking’)or translocation of the protein backbone through a binding cleft(‘threading’), allowing the threaded chain to refoldspontaneously. The observed hydrolysis of ATP would providethe energy for active unfolding. These hypotheses can be appliedto both monomeric folding and oligomeric assembly and are sufficientlydetailed to be open to directed experimental verification.     

An improved prediction of catalytic residues in enzyme structures

The protein databases contain a huge number of function unknown proteins, including many proteins with newly determined 3D structures resulted from the Structural Genomics Projects. To accelerate experiment-based assignment of function, de novo prediction of protein functional sites, like active sites in enzymes, becomes increasingly important. Here, we attempted to improve the prediction of catalytic residues in enzyme structures by seeking and refining different encodings (i.e. residue properties) as well as employing new machine learning algorithms. In particular, considering that catalytic residues can often reveal specific network centrality when representing enzyme structure as a residue contact network, the corresponding measurement (i.e. closeness centrality) was used as one of the most important encodings in our new predictor. Meanwhile, a genetic algorithm integrated neural network (GANN) was also employed. Thanks to the above strategies, our GANN predictor demonstrated a high accuracy of 91.2% in the prediction of catalytic residues based on balanced datasets (i.e. the 1:1 ratio of catalytic to non-catalytic residues). When the GANN method was optimally applied to real enzyme structures, 73.9% of the tested structures had the active site correctly located. Compared with two existing methods, the proposed GANN method also demonstrated a better performance.