NetTurnP--neural network prediction of beta-turns by use of evolutionary information and predicted protein sequence features

β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino acid sequence. The individual β-turn types I, I', II, II', VIII, VIa1, VIa2, VIba and IV have been predicted based on classifications by PROMOTIF, and the two-class prediction of β-turn or not is a superset comprised of all β-turn types. The performance is evaluated using a golden set of non-homologous sequences known as BT426. Our two-class prediction method achieves a performance of: MCC=0.50, Qtotal=82.1%, sensitivity=75.6%, PPV=68.8% and AUC=0.864. We have compared our performance to eleven other prediction methods that obtain Matthews correlation coefficients in the range of 0.17-0.47. For the type specific β-turn predictions, only type I and II can be predicted with reasonable Matthews correlation coefficients, where we obtain performance values of 0.36 and 0.31, respectively. CONCLUSION: The NetTurnP method has been implemented as a webserver, which is freely available at http://www.cbs.dtu.dk/services/NetTurnP/. NetTurnP is the only available webserver that allows submission of multiple sequences.     

Prediction of β-Turns

Kohonen's self-organization model, a neural network model, is applied to predict the β-turns in proteins. There are 455 β-turn tetrapeptides and 3807 non-β-turn tetrapeptides in the training database. The rates of correct prediction for the 110 β-turn tetrapeptides and 30,229 non-β-turn tetrapeptides in the testing database are 81.8% and 90.7%, respectively. The high quality of prediction of neural network model implies that the residue-coupled effect along a polypeptide chain is important for the formation of reversal turns, such as β-turns, during the process of protein folding.     

Classification and Prediction of β-Turn Types

Although a β-turn consists of only four amino acids, it assumes many different types in proteins. Is this basically dependent on the tetrapeptide sequence alone or is it due to a variety of interactions with the other part of a protein? To answer this question, a residue-coupled model is proposed that can reflect the sequence-coupling effect for a tetrapeptide in not only a β-turn or non-β-turn, but also different types of a β-turn. The predicted results by the model for 6022 tetrapeptides indicate that the rates of correct prediction for β-turn types I, I′, II, II′, VI, and VIII and non-β-turns are 68.54%, 93.60%, 85.19%, 97.75%, 100%, 88.75%, and 61.02%, respectively. Each of these seven rates is significantly higher than $\frac{1}{7}$ = 14.29%, the completely randomized rate, implying that the formation of different β-turn types or non-β-turns is considerably correlated with the sequences of a tetrapeptide.     

Prediction of β-turns in proteins by 1-4 and 2-3 correlation model

A 1–4 and 2–3 residue-correlation model is proposed to predict the β-turns in proteins. The average rate of correct prediction for the 455 β-turn tetrapeptides and 4018 non-β-turn tetrapeptides in the training data base is 80.1%, and that for the 223 β-turn tetrapeptides and 12562 non-β-turn tetrapeptides in the testing data base is 80.9%. Compared with the rates of correct prediction based on the residue-independent model reported previously, the quality of prediction is significantly improved by the new model, implying that the correlation effect between the 1st and the 4th residues and that between the 2nd and 3rd residues along a tetrapeptide are important for forming a β-turn in a protein during the process of its folding. © 1997 John Wiley & Sons, Inc. Biopoly 41: 673–702, 1997     

Histone-DNA contacts in the 167 bp 2-turn core particle.

The histone-DNA contacts in the 167 bp 2-turn core particle have been compared with those in the 146 bp 1.75-turn core particle by the methodology developed by Mirzabekov and his colleagues. The contacts in the 167 bp 2-turn core particle retain the essential features of those in the 146 bp 1.75-turn core particle but contacts for histones H3 and H2A were found in the 10 bp extension that discriminates the two particles. In addition the contact for histone H2A near the dyad axis was far more pronounced in the case of the 146 bp core particle.     

Circular dichroism of type 13 β-turn in linear tripeptides containing L-proline and D-alanine

The linear tripeptides tBoc-L-Prolyl-D-alanyl-L-leucine and tBoc-L-prolyl-D-alanyl-L-valine have been shown, from circular dichroism (CD) and infrared spectral data, to take up the 4 → 1 hydrogen bonded β-turn conformation in organic solvents. The CD spectra of these tripeptides in trifluoroethanol exhibit a positive n → π^★ band around 220 nm contrary to the usual negative band observed for the type II β-turn. The observed CD spectra of the tripeptides provide the first examples of those predicted theoretically by Woody for peptides containing L,D sequences and adopting the Venkatachalam type 13 β-turn. This conformation is seen to revert to the type II β-turn when the N-terminal protecting group is acetyl or when the C-terminal residue is glycine. These data are shown to have a direct bearing on the interpretation of the CD spectra of globular proteins.     

Vacuum ultraviolet circular dichroism spectrum of beta-turn in solution.

The vacuum ultraviolet circular dichroism spectrum of an isolated 4 → 1 hydrogen bonded β-turn is reported. The observed spectrum of N-acetyl-Pro-Gly-Leu-OH at ? 40°C in trifluoroethanol is in good agreement with the theoretically calculated CD spectrum of the β-turn conformation. This spectrum, particularly the presence of a strong negative band around 180 nm and a large ratio $\text{[θ]}{}_{201}\text{[θ]}{}_{225}$, can be taken as a characteristic feature of the isolated β-turn conformation. These CD spectral features can thus be used to distinguish the β-turn conformation from the β-structure in solution.     

γ-turn types prediction in proteins using the two-stage hybrid neural discriminant model

Due to the slightly success of protein secondary structure prediction using the various algorithmic and non-algorithmic techniques, similar techniques have been developed for predicting γ-turns in proteins by Kaur and Raghava [2003. A neural-network based method for prediction of γ-turns in proteins from multiple sequence alignment. Protein Sci. 12, 923-929]. However, the major limitation of previous methods was inability in predicting γ-turn types. In a recent investigation we introduced a sequence based predictor model for predicting γ-turn types in proteins [Jahandideh, S., Sabet Sarvestani, A., Abdolmaleki, P., Jahandideh, M., Barfeie, M, 2007a. γ-turn types prediction in proteins using the support vector machines. J. Theor. Biol. 249, 785-790]. In the present work, in order to analyze the effect of sequence and structure in the formation of γ-turn types and predicting γ-turn types in proteins, we applied novel hybrid neural discriminant modeling procedure. As the result, this study clarified the efficiency of using the statistical model preprocessors in determining the effective parameters. Moreover, the optimal structure of neural network can be simplified by a preprocessor in the first stage of hybrid approach, thereby reducing the needed time for neural network training procedure in the second stage and the probability of overfitting occurrence decreased and a high precision and reliability obtained in this way.     

β-Turn structure and intramolecular interaction of tetrapeptides containing Asp and Lys

Nmr and CD studies of terminally protected tetrapeptides were carried out in aqueous and DMSO solutions to investigate the formation and stabilization of the β-turn structure. Boc-Gly-Lys-Asp-Gly-OMe and Boc-Asp-Lys-Asp-Gly-OMe appear to have a tendency to adopt a β-turn structure in aqueous solution from the CD spectra and temperature-dependence studies of the amide proton chemical shifts. The side-chain conformation of the Asp residue depends greatly on its ionization state but was not affected by the deprotonation of the neighboring Lys side chain. There is evidence for an intramolecular interaction between the Asp and Lys side chains of Boc-Gly-Lys-Asp-Gly-OMe. Such an interaction can contribute to the stabilization of the β-turn structure.     

Design,synthesis and evaluation of γ-turn mimetics as LSD1-selective inhibitors

Lysine-specific demethylase 1 (LSD1) is an attractive molecular target for cancer therapy. We have previously reported potent LSD1-selective inhibitors (i.e., NCD18, NCD38, and their analogs) consisting of trans-2-phenylcyclopropylamine (PCPA) or trans-2-arylcyclopropylamine (ACPA) and a lysine moiety that could form a γ-turn structure in the active site of LSD1. Herein we report the design, synthesis and evaluation of γ-turn mimetic compounds for further improvement of LSD1 inhibitory activity and anticancer activity. Among a series of γ-turn mimetic compounds synthesized by a Mitsunobu-reaction-based amination strategy, we identified 1n as a potent and selective LSD1 inhibitor. Compound 1n induced cell cycle arrest and apoptosis through histone methylation in human lung cancer cells. The γ-turn mimetics approach should offer new insights into drug design for LSD1-selective inhibitors.     

Introduction of a new scheme for classifying β-turns in protein structures

Protein β-turn classification remains an area of ongoing development in structural biology research. While the commonly used nomenclature defining type I, type II and type IV β-turns was introduced in the 1970s and 1980s, refinements of β-turn type definitions have been introduced as recently as 2019 by Dunbrack, Jr and co-workers who expanded the number of β-turn types to 18 (Shapovalov et al, PLOS Computat. Biol., 15, e1006844, 2019). Based on their analysis of 13 030 turns from 1074 ultrahigh resolution (≤1.2 Å) protein structures, they used a new clustering algorithm to expand the definitions used to classify protein β-turns and introduced a new nomenclature system. We recently encountered a specific problem when classifying β-turns in crystal structures of pentapeptide repeat proteins (PRPs) determined in our lab that are largely composed of β-turns that often lie close to, but just outside of, canonical β-turn regions. To address this problem, we devised a new scheme that merges the Klyne-Prelog stereochemistry nomenclature and definitions with the Ramachandran plot. The resulting Klyne-Prelog-modified Ramachandran plot scheme defines 1296 distinct potential β-turn classifications that cover all possible protein β-turn space with a nomenclature that indicates the stereochemistry of i + 1 and i + 2 backbone dihedral angles. The utility of the new classification scheme was illustrated by re-classification of the β-turns in all known protein structures in the PRP superfamily and further assessed using a database of 16 657 high-resolution protein structures (≤1.5 Å) from which 522 776 β-turns were identified and classified.     

Using predicted shape string to enhance the accuracy of <Emphasis Type="Bold">γ</Emphasis>-turn prediction

Numerous methods for predicting γ-turns in proteins have been developed. However, the results they generally provided are not very good, with a Matthews correlation coefficient (MCC) ≤0.18. Here, an attempt has been made to develop a method to improve the accuracy of γ-turn prediction. First, we employ the geometric mean metric as optimal criterion to evaluate the performance of support vector machine for the highly imbalanced γ-turn dataset. This metric tries to maximize both the sensitivity and the specificity while keeping them balanced. Second, a predictor to generate protein shape string by structure alignment against the protein structure database has been designed and the predicted shape string is introduced as new variable for γ-turn prediction. Based on this perception, we have developed a new method for γ-turn prediction. After training and testing the benchmark dataset of 320 non-homologous protein chains using a fivefold cross-validation technique, the present method achieves excellent performance. The overall prediction accuracy Q _total can achieve 92.2% and the MCC is 0.38, which outperform the existing γ-turn prediction methods. Our results indicate that the protein shape string is useful for predicting protein tight turns and it is reasonable to use the dihedral angle information as a variable for machine learning to predict protein folding. The dataset used in this work and the software to generate predicted shape string from structure database can be obtained from anonymous ftp site freely.     

Proline-containing β-turns in peptides and proteins. II. Physicochemical studies on tripeptides with the Pro-Gly sequence

Amino acids are known to differ in their individual preferences for each of the four positions of the β-turn conformation formed by tetrapeptide segments. Proline and glycine show relatively high preferences for positions 2 and 3, respectively, of the β-turn. Using tripeptides of the type N-acetyl-Pro-Gly-X-OH, where X = Gly, Ala, Leu, Ile, and Phe, we have sought to study the influence of the 4th residue X on the stability of the β-turn conformation in these tripeptides. Our nmr and CD results show that the β-turn stability is quite significantly governed by the nature of the amino acid residue at this position in the following order: Leu > Ala > Ile, Gly > Phe.     

Structural and functional importance of theβ-turn in proteins. Studies on proline-containing peptides

We report here two sets of results on proline-containing linear peptides, one of which brings out the role of theβ-turn conformation in the structure of nascent collagen while the other points to the functional importance of the β-turn in calcium-binding proteins. Based on the data on peptides containing the -Pro-Gly-sequence, we had proposed and experimentally verified that theβ-turn conformation in these peptides is a structural requirement for the enzymic hydroxylation of the proline residues in the nascent (unhydroxylated) procollagen molecule. Our recent data, presented here, on the conformation of peptides containing both the -Pro-Gly- and -Gly-Pro-sequences reveal that while theβ-turn in the substrate molecule is required at the catalytic site of prolyl hydroxylase, the polyproline-II structure is necessary for effective binding at the active site of the enzyme. Thus, peptides containing either theβ-turn or the polyproline-II structure alone are found to act only as inhibitors while those with the polyproline-II followed byβ-turn serve as substrates of the enzyme. In another study, we have synthesized the two linear peptides: Boc-Pro-D-Ala-Ala-NHCH₃ and Boc-Pro-Gly-Ala-NHCH₃ each of which adopts, in solution, a structure with two consecutiveβ-turns, as judged from circular dichroism, infrared and nuclear magnetic resonance data. Drastic spectral changes are seen in these peptides on binding to Ca²⁺. Both the peptides show a distinct specificity to Ca²⁺ over Mg²⁺, Na⁺ and Li₊. A conformational change in the peptides occurs on Ca²⁺ binding which brings together the carbonyl groups to coordinate with the metal ion. These results imply a functional role for theβ-turn in Ca²⁺ — binding proteins.     

pi-Turns: types,systematics and the context of their occurrence in protein structures

Background  

For a proper understanding of protein structure and folding it is important to know if a polypeptide segment adopts a conformation inherent in the sequence or it depends on the context of its flanking secondary structures. Turns of various lengths have been studied and characterized starting from three-residue γ-turn to six-residue π-turn. The Schellman motif occurring at the C-terminal end of α-helices is a classical example of hydrogen bonded π-turn involving residues at (i) and (i+5) positions. Hydrogen bonded and non-hydrogen bonded β- and α-turns have been identified previously; likewise, a systematic characterization of π-turns would provide valuable insight into turn structures.     

Energy minimization studies on α-turns

Using a grid search technique, the entire conformational space of a system of four linked peptide units (tetrapeptide) was scanned to pick out geometrically possible 5→1 type hydrogen-bonded conformations defined as an α-turn. The energy minimization of these conformations led to 23 distinct minimum energy conformations (MECs) falling in 13 different classes. The presence of β and γ turn type hydrogen bonds along with 5→1 type hydrogen bond gave conformational variability in a given class. The occurrence of bifurcated hydrogen bonding network was a characteristic feature of most of the MECs. In many prototype MECs non-glycyl residues such as Ala and Pro could be accommodated. Comparison of MECs with the α-turn examples that are observed in proteins showed that the conformationally worked out MECs occurred in isolation in proteins, with the α-helical α-turn being distinctly the most predominant. © 1998 European Peptide Society and John Wiley & Sons, Ltd.     

Nmr and molecular modeling investigations of the neuropeptide bradykinin in three different solvent systems: DMSO, 9 : 1 dioxane/water,and in the presence of 7.4 mM lyso phosphatidylcholine micelles

The linear nonapeptide hormone bradykinin (Arg¹-Pro²-Pro³-Gly⁴-Phe⁵-Ser⁶-Pro⁷-Phe⁸-Arg⁹) is involved, either directly or indirectly, in a wide variety of physiological processes, particularly pain and hyperanalgesia. Additional evidence suggests that bradykinin also plays a major role in inflammatory response, asthma, sepsis, and symptoms associated with the rhinoviral infection. It has long been speculated that a β-turn at the C-terminus of bradykinin plays a major role in the biological activity of the neuropeptide. The β-turn forming potential of bradykinin in three vastly different local chemical environments, DMSO, 9 : 1 dioxane/water, and in the presence of 7.4 mM lyso phosphatidylcholine micelles, was investigated using two-dimensional homonuclear nmr experiments coupled with simulated annealing calculations. The results of these investigations show that in all three systems residues 6–9 of the C-terminus adopt very similar β-turn like structures. These results suggest that the β-turn at the C-terminus of bradykinin is an important secondary structural feature for receptor recognition and binding. © 1994 John Wiley & Sons, Inc.     

Prediction of β-Turns

Kohonen's self-organization model, a neural network model, is applied to predict the -turns in proteins. There are 455 -turn tetrapeptides and 3807 non--turn tetrapeptides in the training database. The rates of correct prediction for the 110 -turn tetrapeptides and 30,229 non--turn tetrapeptides in the testing database are 81.8% and 90.7%, respectively. The high quality of prediction of neural network model implies that the residue-coupled effect along a polypeptide chain is important for the formation of reversal turns, such as -turns, during the process of protein folding.     

Conformational analysis of a snake neurotoxin by prediction from sequence, circular dichroism, and raman spectroscopy.

The secondary structure of the major neurotoxin from the sea snake Lapemis hardwickii was investigated by several methods of conformational analysis: structure prediction, circular dichroism, and laser Raman spectroscopy. From the primary structure, secondary structure prediction yielded two regions of β-sheet structure at residues 1–7 and 41–45. β-Turns were predicted at residues 14–17, 20–23, 30–33, 37–40, and 46–49. From the predictions, the toxin appears to be composed of approximately 20% β-sheet and 33% β-turn. The CD spectrum of the native toxin appears to be a hybrid of model spectra for β-sheet and β-turn proteins. The pH perturbation studies on the toxin observed by CD demonstrated that the toxin is a very stable molecule except at extremely high or low pH values. The Raman data indicated that the toxin contains both antiparallel β-sheet and β-turn structure. Using two methods of secondary structure quantitation from Raman spectra the molecule was calculated to contain 35% β-sheet from one method and 27% from the other. Overall, the various methods demonstrate that the toxin is composed of β-sheet and β-turn structure with little or no α-helix present. From the comparison of these different techniques appreciation can be gained for the necessity of several methods when identifying and quantitating secondary structure.