The influence of the local sequence environment on RNA loop structures

RNA folding is assumed to be a hierarchical process. The secondary structure of an RNA molecule, signified by base-pairing and stacking interactions between the paired bases, is formed first. Subsequently, the RNA molecule adopts an energetically favorable three-dimensional conformation in the structural space determined mainly by the rotational degrees of freedom associated with the backbone of regions of unpaired nucleotides (loops). To what extent the backbone conformation of RNA loops also results from interactions within the local sequence context or rather follows global optimization constraints alone has not been addressed yet. Because the majority of base stacking interactions are exerted locally, a critical influence of local sequence on local structure appears plausible. Thus, local loop structure ought to be predictable, at least in part, from the local sequence context alone. To test this hypothesis, we used Random Forests on a nonredundant data set of unpaired nucleotides extracted from 97 X-ray structures from the Protein Data Bank (PDB) to predict discrete backbone angle conformations given by the discretized η/θ-pseudo-torsional space. Predictions on balanced sets with four to six conformational classes using local sequence information yielded average accuracies of up to 55%, thus significantly better than expected by chance (17%-25%). Bases close to the central nucleotide appear to be most tightly linked to its conformation. Our results suggest that RNA loop structure does not only depend on long-range base-pairing interactions; instead, it appears that local sequence context exerts a significant influence on the formation of the local loop structure.     

Protein local conformations arise from a mixture of Gaussian distributions

The classical approaches for protein structure prediction rely either on homology of the protein sequence with a template structure or on ab initio calculations for energy minimization. These methods suffer from disadvantages such as the lack of availability of homologous template structures or intractably large conformational search space, respectively. The recently proposed fragment library based approaches first predict the local structures,which can be used in conjunction with the classical approaches of protein structure prediction. The accuracy of the predictions is dependent on the quality of the fragment library. In this work, we have constructed a library of local conformation classes purely based on geometric similarity. The local conformations are represented using Geometric Invariants, properties that remain unchanged under transformations such as translation and rotation, followed by dimension reduction via principal component analysis. The local conformations are then modeled as a mixture of Gaussian probability distribution functions (PDF). Each one of the Gaussian PDF's corresponds to a conformational class with the centroid representing the average structure of that class. We find 46 classes when we use an octapeptide as a unit of local conformation. The protein 3-D structure can now be described as a sequence of local conformational classes. Further, it was of interest to see whether the local conformations can be predicted from the amino acid sequences. To that end,we have analyzed the correlation between sequence features and the conformational classes.     

Protein local conformations arise from a mixture of Gaussian distributions

The classical approaches for protein structure prediction rely either on homology of the protein sequence with a template structure or on ab initio calculations for energy minimization. These methods suffer from disadvantages such as the lack of availability of homologous template structures or intractably large conformational search space, respectively. The recently proposed fragment library based approaches first predict the local structures, which can be used in conjunction with the classical approaches of protein structure prediction. The accuracy of the predictions is dependent on the quality of the fragment library. In this work, we have constructed a library of local conformation classes purely based on geometric similarity. The local conformations are represented using Geometric Invariants, properties that remain unchanged under transformations such as translation and rotation, followed by dimension reduction via principal component analysis. The local conformations are then modeled as a mixture of Gaussian probability distribution functions (PDF). Each one of the Gaussian PDF’s corresponds to a conformational class with the centroid representing the average structure of that class. We find 46 classes when we use an octapeptide as a unit of local conformation. The protein 3-D structure can now be described as a sequence of local conformational classes. Further, it was of interest to see whether the local conformations can be predicted from the amino acid sequences. To that end, we have analyzed the correlation between sequence features and the conformational classes.     

Optimized representations and maximal information in proteins

In an effort to quantify loss of information in the processing of protein bioinformatic data, we examine how representations of amino acid sequence and backbone conformation affect the quantity of accessible structural information from local sequence. We propose a method to extract the maximum amount of peptide backbone structural information available in local sequence fragments, given a finite structural data set. Using methods of information theory, we develop an unbiased measure of local structural information that gauges changes in structural distributions when different representations of secondary structure and local sequence are used. We find that the manner in which backbone structure is represented affects the amount and quality of structural information that may be extracted from local sequence. Representations based on virtual bonds capture more structural information from local sequence than a three-state assignment scheme (helix/strand/loop). Furthermore, we find that amino acids show significant kinship with respect to the backbone structural information they carry, so that a collapse of the amino acid alphabet can be accomplished without severely affecting the amount of extractable information. This strategy is critical in optimizing the utility of a limited database of experimentally solved protein structures. Finally, we discuss the similarities within and differences between groups of amino acids in their roles in the local folding code and recognize specific amino acids critical in the formation of local structure.     

The crystal structure of d(CCCCGGGG): a new A-form variant with an extended backbone conformation

The crystal structure of d(CCCCGGGG) has been determined at a resolution of 2.25 A. The oligomers crystallize as A-DNA duplexes occupying crystallographic two-fold axes. The backbone conformation is, in general, similar to that observed in previously reported crystal structures of A-DNA fragments, except for the central linkage, where it adopts an extended structure resulting from all trans conformation at the P-O5'-C5'-C4' bonds. This type of conformation facilitates interstrand stacking between the guanines at the C-G site. The local helix twist at this step is very small (25 degrees) compared to an overall average of 33.5 degrees. The unique structure of the C-G base-pair step, namely the extended backbone and the distinct stacking geometry, may be an important feature in the recognition mechanism between double-stranded DNA molecules and restriction endonucleases such as Msp I, which cuts the sequence CCGG very specifically with a rate unaffected by neighboring base pairs.     

Potential energy function for continuous state models of globular proteins.

One of the approaches to protein structure prediction is to obtain energy functions which can recognize the native conformation of a given sequence among a zoo of conformations. The discriminations can be done by assigning the lowest energy to the native conformation, with the guarantee that the native is in the zoo. Well-adjusted functions, then, can be used in the search for other (near-) natives. Here the aim is the discrimination at relatively high resolution (RMSD difference between the native and the closest nonnative is around 1 A) by pairwise energy potentials. The potential is trained using the experimentally determined native conformation of only one protein, instead of the usual large survey over many proteins. The novel feature is that the native structure is compared to a vastly wider and more challenging array of nonnative structures found not only by the usual threading procedure, but by wide-ranging local minimization of the potential. Because of this extremely demanding search, the native is very close to the apparent global minimum of the potential function. The global minimum property holds up for one other protein having 60% sequence identity, but its performance on completely dissimilar proteins is of course much weaker.     

The solution structure of a chimeric LEKTI domain reveals a chameleon sequence

The conversion of an alpha-helical to a beta-strand conformation and the presence of chameleon sequences are fascinating from the perspective that such structural features are implicated in the induction of amyloid-related fatal diseases. In this study, we have determined the solution structure of a chimeric domain (Dom1PI) from the multidomain Kazal-type serine proteinase inhibitor LEKTI using multidimensional NMR spectroscopy. This chimeric protein was constructed to investigate the reasons for differences in the folds of the homologous LEKTI domains 1 and 6 [Lauber, T., et al. (2003) J. Mol. Biol. 328, 205-219]. In Dom1PI, two adjacent phenylalanine residues (F28 and F29) of domain 1 were substituted with proline and isoleucine, respectively, as found in the corresponding P4' and P5' positions of domain 6. The three-dimensional structure of Dom1PI is significantly different from the structure of domain 1 and closely resembles the structure of domain 6, despite the sequence being identical to that of domain 1 except for the two substituted phenylalanine residues and being only 31% identical to the sequence of domain 6. The mutation converted a short 3(10)-helix into an extended loop conformation and parts of the long COOH-terminal alpha-helix of domain 1 into a beta-hairpin structure. The latter conformational change occurs in a sequence stretch distinct from the region containing the substituted residues. Therefore, this switch from an alpha-helical structure to a beta-hairpin structure indicates a chameleon sequence of seven residues. We conclude that the secondary structure of Dom1PI is determined not only by the local protein sequence but also by nonlocal interactions.     

A new method for building protein conformations from sequence alignments with homologues of known structure

We describe a largely automatic procedure for building protein structures from sequence alignments with homologues of known structure. This procedure uses simple rules by which multiple sequence alignments can be translated into distance and chirality constraints, which are then used as input for distance geometry calculations. By this means one obtains an ensemble of conformations for the unknown structure that are compatible with the rules employed, and the differences among these conformations provide an indication of the reliability of the structure prediction. The overall approach is demonstrated here by applying it to several Kazal-type trypsin inhibitors, for which experimentally determined structures are available. On the basis of our experience with these test problems, we have further predicted the conformation of the human pancreatic secretory trypsin inhibitor, for which no experimentally determined structure is presently available.     

Structure and folding of four putative kink turns identified in structured RNA species in a test of structural prediction rules

k-Turns are widespread key architectural elements that occur in many classes of RNA molecules. We have shown previously that their folding properties (whether or not they fold into their tightly kinked structure on addition of metal ions) and conformation depend on their local sequence, and we have elucidated a series of rules for prediction of these properties from sequence. In this work, we have expanded the rules for prediction of folding properties, and then applied the full set to predict the folding and conformation of four probable k-turns we have identified amongst 224 structured RNA species found in bacterial intergenenic regions by the Breaker lab (1). We have analyzed the ion-dependence of folding of the four k-turns using fluorescence resonance energy transfer, and determined the conformation of two of them using X-ray crystallography. We find that the experimental data fully conform to both the predicted folding and conformational properties. We conclude that our folding rules are robust, and can be applied to new k-turns of unknown characteristics with confidence.     

The Crystal Structure of d(CCCCGGGG): A New A-Form Variant with an Extended Backbone Conformation

Abstract

The crystal structure of d(CCCCGGGG) has been determined at a resolution of 2.25Å. The oligomers crystallize as A-DNA duplexes occupying crystallographic two-fold axes. The backbone conformation is, in general, similar to that observed in previously reported crystal structures of A-DNA fragments, except for the central linkage, where it adopts an extended structure resulting from all trans conformation at the P-05′-C5′-C4′ bonds. This type of conformation facilitates interstrand stacking between the guanines at the C-G site. The local helix twist at this step is very small (25°) compared to an overall average of 33.5°. The unique structure of the C-G base-pair step, namely the extended backbone and the distinct stacking geometry, may be an important feature in the recognition mechanism between double- stranded DNA molecules and restriction endonucleases such as Msp I, which cuts the sequence CCGG very specifically with a rate unaffected by neighboring base pairs.     

Conservation of cis prolyl bonds in proteins during evolution

In proteins and peptides, the vast majority of peptide bonds occurs in trans conformation, but a considerable fraction (about 5%) of X-Pro bonds adopts the cis conformation. Here we study the conservation of cis prolyl residues in evolutionary related proteins. We find that overall, in contrast to local, protein sequence similarity is a clear indicator for the conformation of prolyl residues. We observe that cis prolyl residues are more often conserved than trans prolyl residues, and both are more conserved than the surrounding amino acids, which show the same extent of conservation as the whole protein. The pattern of amino acid exchanges differs between cis and trans prolyl residues. Also, the cis prolyl bond is maintained in proteins with sequence identity as low as 20%. This finding emphasizes the importance of cis peptide bonds in protein structure and function.     

Exploring local and non-local interactions for protein stability by structural motif engineering

In order to probe the relative contribution of local and non-local interactions to the thermodynamic stability of proteins, we have devised an experimental approach based on a combination of motif engineering and sequence shuffling. Candidate chain segments in an immunoglobulin V(L) domain were identified whose conformation is proposed to be dominated by non-local interactions. Locally interacting structural motifs of a different conformation were then constructed as replacements, by introducing motif consensus sequences. We find that all nine replacements we constructed systematically reduce the folding cooperativity. By comparing this destabilising effect with the folding transitions of shuffled sequences for three of these motifs, we estimate the contribution of local, native interactions to the free energy of folding. Our results suggest that local and non-local interactions contribute to stability by an approximately equal amount, but that local interactions stabilise by increasing the resistance to denaturation while non-local interactions increase folding cooperativity. The systematic loss of stability by sequence shuffling in these host-guest experiments suggests that the designed interactions indeed are present in the native state, thus consensus sequence engineering may be a useful tool in structure design, but non-local interactions must be taken into account for global stability engineering. Statistical approaches are powerful tools for engineering protein structure and stability, but an analysis based on local sequence propensities alone does not adequately represent the balance of sequence and context in protein structures.     

Base sequence, local helix structure, and macroscopic curvature of A-DNA and B-DNA

Given a specified DNA sequence and starting with an idealized conformation for the double helix (A-DNA or B-DNA), the dependence of conformational energy on variations in the local geometry of the double helix can be examined by computer modeling. By averaging over all thermally accessible states, it is possible to determine 1) how the optimum local structure differs from the initial idealized conformation and 2) the energetic costs of small structural deformations. This paper describes such a study. Tables are presented for the prediction of helix twist angles and base pair roll angles for both A-DNA and B-DNA when the sequence has been specified. Local deviations of helix parameters from their average values can accumulate to produce a net curvature of the molecule, a curvature that can be sharp enough to be experimentally detectable. As an independent check on the method, the calculations provide predictions for the longitudinal compressibility (Young's modulus) and the average torsional stiffness, both of which are in good agreement with experimental values. In examining the role of sequence-dependent variations in helix structure for the recognition of specific sequences by proteins, we have calculated the energy needed to deform the self-complementary hexanucleotide d(CAATTG) to match the local geometry of d(GAATTC), which is the sequence recognized by the EcoRI restriction endonuclease. That energy would be sufficient to reduce the binding of the incorrect sequence to the protein by over 2 orders of magnitude relative to the correct sequence.     

Local protein sequence similarity does not imply a structural relationship

A database search often will find a seemingly strong sequence similarity between two fragments of proteins that are not expected to have an evolutionary or functional relationship. It is tempting to suggest that the two fragments will adopt a similar conformation due to a common pattern of residues that dictate a particular substructure. To investigate the likelihood of such a structural similarity, local sequence similarities between proteins of known conformation were identified by a standard database search algorithm. Significant sequence similarity was identified as when the chance probability of obtaining the relatedness score from a scan of the entire database was less than 1%. In this region both true homologies and false homologies are detected. A total of 69 false homologies was located of length between 20 and 262 aligned positions. Many of these alignments had approximately 25% sequence identity and a further 25% of conservative changes. However, the results show in general these aligned fragments did not have a significant similarity in secondary or tertiary structure. Thus local sequence does not indicate a structural similarity when there is neither an evolutionary nor functional explanation to support this. Accordingly structure predictions based on finding a local sequence similarity with an evolutionary unrelated protein of known conformation are unlikely to be valid.     

Assembly of polypeptide and protein backbone conformations from low energy ensembles of short fragments: development of strategies and construction of models for myoglobin, lysozyme, and thymosin beta 4.

Recently we developed methods for the construction of knowledge-based mean fields from a data base of known protein structures. As shown previously, this approach can be used to calculate ensembles of probable conformations for short fragments of polypeptide chains. Here we develop procedures for the assembly of short fragments to complete three-dimensional models of polypeptide chains. The amino acid sequence of a given protein is decomposed into all possible overlapping fragments of a given length, and an ensemble of probable conformations is calculated for each fragment. The fragments are assembled to a complete model by choosing appropriate conformations from the individual ensembles and by averaging over equivalent angles. Finally a consistent model is obtained by rebuilding the conformation from the average angles. From the average angles the local variability of the structure can be calculated, which is a useful criterion for the reliability of the model. The procedure is applied to the calculation of the local backbone conformations of myoglobin and lysozyme whose structures have been solved by X-ray analysis and thymosin beta 4, a polypeptide of 43 amino acid residues whose structure was recently investigated by NMR spectroscopy. We demonstrate that substantial fractions of the calculated local backbone conformations are similar to the experimentally determined structures.     

Molecular flexibility in protein-DNA interactions

In living cells protein-DNA interactions are fundamental processes. Here, we compare the 3D structures of several DNA-binding proteins frequently determined with and without attached DNA. We studied the global structure (backbone-traces) as well as the local structure (binding sites) by comparing pair-wise the related atoms. The DNA-interaction sites of uncomplexed proteins show conspicuously high local structural flexibility. Binding to DNA results in specific local conformations, which are clearly distinct from the unbound states. The adaptation of the protein's binding site to DNA can never be described by the lock and key model but in all cases by the induced fit model. Conformational changes in the seven protein backbone traces take place in different ways. Two of them dock onto DNA without a significant change, while the other five proteins are characterized by a backbone conformation change caused by DNA docking. In the case of three proteins of the latter group the DNA-complexed conformation also occurs in a few uncomplexed structures. This behavior can be described by a conformational ensemble, which is narrowed down by DNA docking until only one single DNA-complexed conformation occurs. Different docking models are discussed and each of the seven proteins is assigned to one of them.     

Sequence dependent DNA conformations: Raman spectroscopic studies and a model of action of restriction enzymes

Raman spectra have been examined to clarify the polymorphic forms of DNA, A, B, and Z forms. From an analysis we found that the guanine ring breathing vibration is sensitive to its local conformation. Examination of nine crystals of guanosine residues in which the local conformations are well established revealed that a guanosine residue with a C3′endo-anti gives a strong line at 666±2cm⁻¹, O4′endo-anti at 682 cm⁻¹, Cl′exo-anti at 673 cm⁻¹, C2′endo-anti at 677 cm⁻¹ and syn-forms around 625 cm⁻¹. Using this characteristic line, we were able to obtain the local conformations of guanosine moieties in poly(dG-dC).Such a sequence derived variation is suggested to be recognized by sequence specific proteins such as restriction enzymes. We found a correlation between sequence dependent DNA conformation and a mode of action of restriction enzymes. The cutting mode of restriction enzymes is classified into three groups. The classification of whether the products have blunt ends, two-base-long cohesive ends, or four-base-long cohesive ends depends primarily on the substrate, not on the enzyme. It is suggested that sequence dependent DNA conformation causes such a classification by the use of the Calladine-Dickerson analysis. In the recognition of restriction enzymes, the methyl group in a certain sequence is considered to play an important role by changing the local conformation of DNA.     

Computation of conformational entropy from protein sequences using the machine-learning method--application to the study of the relationship between structural conservation and local structural stability

A complete protein sequence can usually determine a unique conformation; however, the situation is different for shorter subsequences--some of them are able to adopt unique conformations, independent of context; while others assume diverse conformations in different contexts. The conformations of subsequences are determined by the interplay between local and nonlocal interactions. A quantitative measure of such structural conservation or variability will be useful in the understanding of the sequence-structure relationship. In this report, we developed an approach using the support vector machine method to compute the conformational variability directly from sequences, which is referred to as the sequence structural entropy. As a practical application, we studied the relationship between sequence structural entropy and the hydrogen exchange for a set of well-studied proteins. We found that the slowest exchange cores usually comprise amino acids of the lowest sequence structural entropy. Our results indicate that structural conservation is closely related to the local structural stability. This relationship may have interesting implications in the protein folding processes, and may be useful in the study of the sequence-structure relationship.     

Formation of the local secondary structure of proteins: local sequence or environment

An analysis of the occurrence of tetrapeptides in 35 globular proteins for alpha-helix, beta-structure and coil was performed. We concluded that: the conformation of a short polypeptide segment cannot be determined on the basis of the knowledge of the amino acid sequence only; local structures of a protein are formed as the result of interactions within the whole structural domain of the protein as well as interactions with the environment.