Amino acid side chain parameters for correlation studies in biology and pharmacology

Predicting host tropism of influenza A virus proteins using random forest

Background

Majority of influenza A viruses reside and circulate among animal populations, seldom infecting humans due to host range restriction. Yet when some avian strains do acquire the ability to overcome species barrier, they might become adapted to humans, replicating efficiently and causing diseases, leading to potential pandemic. With the huge influenza A virus reservoir in wild birds, it is a cause for concern when a new influenza strain emerges with the ability to cross host species barrier, as shown in light of the recent H7N9 outbreak in China. Several influenza proteins have been shown to be major determinants in host tropism. Further understanding and determining host tropism would be important in identifying zoonotic influenza virus strains capable of crossing species barrier and infecting humans.

Results

In this study, computational models for 11 influenza proteins have been constructed using the machine learning algorithm random forest for prediction of host tropism. The prediction models were trained on influenza protein sequences isolated from both avian and human samples, which were transformed into amino acid physicochemical properties feature vectors. The results were highly accurate prediction models (ACC>96.57; AUC>0.980; MCC>0.916) capable of determining host tropism of individual influenza proteins. In addition, features from all 11 proteins were used to construct a combined model to predict host tropism of influenza virus strains. This would help assess a novel influenza strain's host range capability.

Conclusions

From the prediction models constructed, all achieved high prediction performance, indicating clear distinctions in both avian and human proteins. When used together as a host tropism prediction system, zoonotic strains could potentially be identified based on different protein prediction results. Understanding and predicting host tropism of influenza proteins lay an important foundation for future work in constructing computation models capable of directly predicting interspecies transmission of influenza viruses. The models are available for prediction at http://fluleap.bic.nus.edu.sg.

Polar profile of antiviral peptides from AVPpred Database

Diseases of viral origin in humans are among the most serious threats to health and the global economy. As recent history has shown the virus has a high pandemic potential, among other reasons, due to its ability to spread by air, hence the identification, investigation, containment, and treatment of viral diseases should be considered of paramount importance. In this sense, the bioinformatics research has focused on finding fast and efficient algorithms that can identify highly toxic antiviral peptides and to serve as a first filter, so that trials in the laboratory are substantially reduced. The work presented here contributes to this effort through the use of an algorithm already published by this team, called polarity index method, which identifies with high efficiency antiviral peptides from the exhaustive analysis of the polar profile, using the linear sequence of the peptide. The test carried out included all peptides in APD2 Database and 60 antiviral peptides identified by Kumar and co-workers (Nucleic Acids Res 40:W199-204, 2012), to build its AVPpred algorithm. The validity of the method was focused on its discriminating capacity so we included the 15 sub-classifications of both Databases.

Analysing the substrate multispecificity of a proton-coupled oligopeptide transporter using a dipeptide library

Peptide uptake systems that involve members of the proton-coupled oligopeptide transporter (POT) family are conserved across all organisms. POT proteins have characteristic substrate multispecificity, with which one transporter can recognize as many as 8,400 types of di/tripeptides and certain peptide-like drugs. Here we characterize the substrate multispecificity of Ptr2p, a major peptide transporter of Saccharomyces cerevisiae, using a dipeptide library. The affinities (K_i) of di/tripeptides toward Ptr2p show a wide distribution range from 48 mM to 0.020 mM. This substrate multispecificity indicates that POT family members have an important role in the preferential uptake of vital amino acids. In addition, we successfully establish high performance ligand affinity prediction models (97% accuracy) using our comprehensive dipeptide screening data in conjunction with simple property indices for describing ligand molecules. Our results provide an important clue to the development of highly absorbable peptides and their derivatives including peptide-like drugs.

Proton-coupled oligopeptide transporters (POTs) can recognize and mediate the uptake of up to 8,400 di/tripeptides or peptide-like drugs. Ito et al. comprehensively map the substrate specificity of the yeast POT Ptr2p, and use this information to construct models for the prediction of ligand affinity.

Identification of a binding element for the cytoplasmic regulator FROUNT in the membrane-proximal C-terminal region of chemokine receptors CCR2 and CCR5

Chemokine receptors mediate the migration of leucocytes during inflammation. The cytoplasmic protein FROUNT binds to chemokine receptors CCR2 [chemokine (C-C motif) receptor 2] and CCR5, and amplifies chemotactic signals in leucocytes. Although the interaction between FROUNT and chemokine receptors is important for accurate chemotaxis, the interaction mechanism has not been elucidated. In the present study we identified a 16-amino-acid sequence responsible for high-affinity binding of FROUNT at the membrane-proximal C-terminal intracellular region of CCR2 (CCR2 Pro-C) by yeast two-hybrid analysis. Synthesized peptides corresponding to the CCR2 Pro-C sequence directly interacted with FROUNT in vitro. CCR2 Pro-C was predicted to form an amphipathic helix structure. Residues on the hydrophobic side are completely conserved among FROUNT-binding receptors, suggesting that the hydrophobic side is the responsible element for FROUNT binding. The L316T mutation to the hydrophobic side of the predicted helix decreased the affinity for FROUNT. Co-immunoprecipitation assays revealed that the CCR2 L316T mutation diminished the interaction between FROUNT and full-length CCR2 in cells. Furthermore, this mutation impaired the ability of the receptor to mediate chemotaxis. These findings provide the first description of the functional binding element in helix 8 of CCR2 for the cytosolic regulator FROUNT that mediates chemotactic signalling.

Prediction of protein-protein interaction with pairwise kernel support vector machine

Protein–protein interactions (PPIs) play a key role in many cellular processes. Unfortunately, the experimental methods currently used to identify PPIs are both time-consuming and expensive. These obstacles could be overcome by developing computational approaches to predict PPIs. Here, we report two methods of amino acids feature extraction: (i) distance frequency with PCA reducing the dimension (DFPCA) and (ii) amino acid index distribution (AAID) representing the protein sequences. In order to obtain the most robust and reliable results for PPI prediction, pairwise kernel function and support vector machines (SVM) were employed to avoid the concatenation order of two feature vectors generated with two proteins. The highest prediction accuracies of AAID and DFPCA were 94% and 93.96%, respectively, using the 10 CV test, and the results of pairwise radial basis kernel function are considerably improved over those based on radial basis kernel function. Overall, the PPI prediction tool, termed PPI-PKSVM, which is freely available at http://159.226.118.31/PPI/index.html, promises to become useful in such areas as bio-analysis and drug development.

Construction of protein dendrograms based on amino acid indices and Discrete Fourier Transform

From the literature, existing methods use pairwise percent identity to identify the percentage of similarity between two protein sequences, in order to create a dendrogram. As this is a parametric method of measuring the similarities between proteins, and different parameter may yield different results, this method does not guarantee that the global optimal similarity values will be found. As protein dendrogram construction is used in other areas, such as multiple protein sequence alignments, it is very important that the most related protein sequences to be identified and align first. Furthermore, by using the pairwise percent identity of the protein sequences to construct the dendrograms, the physical characteristics of protein sequences and amino acids are not considered. In this paper, a new method was proposed for constructing protein sequence dendrograms. For this method, Discrete Fourier Transform, was used to construct the distance matrix in combination with the multiple amino acid indices that were used to encode protein sequences into numerical sequences. In order to show the applicability and robustness of the proposed method, a case study was presented by using nine Cluster of Differentiation 4 protein sequences extracted from the UniProt online database.

Probabilistic grammatical model for helix-helix contact site classification

Background

Hidden Markov Models power many state‐of‐the‐art tools in the field of protein bioinformatics. While excelling in their tasks, these methods of protein analysis do not convey directly information on medium‐ and long‐range residue‐residue interactions. This requires an expressive power of at least context‐free grammars. However, application of more powerful grammar formalisms to protein analysis has been surprisingly limited.

Results

In this work, we present a probabilistic grammatical framework for problem‐specific protein languages and apply it to classification of transmembrane helix‐helix pairs configurations. The core of the model consists of a probabilistic context‐free grammar, automatically inferred by a genetic algorithm from only a generic set of expert‐based rules and positive training samples. The model was applied to produce sequence based descriptors of four classes of transmembrane helix‐helix contact site configurations. The highest performance of the classifiers reached AUCROC of 0.70. The analysis of grammar parse trees revealed the ability of representing structural features of helix‐helix contact sites.

Conclusions

We demonstrated that our probabilistic context‐free framework for analysis of protein sequences outperforms the state of the art in the task of helix‐helix contact site classification. However, this is achieved without necessarily requiring modeling long range dependencies between interacting residues. A significant feature of our approach is that grammar rules and parse trees are human‐readable. Thus they could provide biologically meaningful information for molecular biologists.

Amino acid changes in disease-associated variants differ radically from variants observed in the 1000 genomes project dataset

The 1000 Genomes Project data provides a natural background dataset for amino acid germline mutations in humans. Since the direction of mutation is known, the amino acid exchange matrix generated from the observed nucleotide variants is asymmetric and the mutabilities of the different amino acids are very different. These differences predominantly reflect preferences for nucleotide mutations in the DNA (especially the high mutation rate of the CpG dinucleotide, which makes arginine mutability very much higher than other amino acids) rather than selection imposed by protein structure constraints, although there is evidence for the latter as well. The variants occur predominantly on the surface of proteins (82%), with a slight preference for sites which are more exposed and less well conserved than random. Mutations to functional residues occur about half as often as expected by chance. The disease-associated amino acid variant distributions in OMIM are radically different from those expected on the basis of the 1000 Genomes dataset. The disease-associated variants preferentially occur in more conserved sites, compared to 1000 Genomes mutations. Many of the amino acid exchange profiles appear to exhibit an anti-correlation, with common exchanges in one dataset being rare in the other. Disease-associated variants exhibit more extreme differences in amino acid size and hydrophobicity. More modelling of the mutational processes at the nucleotide level is needed, but these observations should contribute to an improved prediction of the effects of specific variants in humans.

In this paper we compare the differences between ‘natural’ and disease-associated amino acid variants at both sequence as well as structural levels. We used data from the 1000 Genomes Project (1 kG), the OMIM database and UniProtKB Humsavar. The results highlight the complex interplay of features from the level of the DNA up to protein sequence and structure. The codon CpG dinucleotide content plays a large role in determining which amino acids mutate. This in turn affects the mutability of amino acids and a clear difference was seen between non-disease and disease variants where amino acids that are naturally very mutable show the opposite trend in the disease-associated data. The current results show evidence for some selection, mainly in that the variants occur slightly more often on the surface of the protein and are much less likely to be annotated as functional than expected by chance. However we should note that even the best definition of functional, taken from structural data, is limited. Even with these caveats, it is clear that the 1 kG variants eschew functional residues as defined here, a trend which is surprisingly even stronger in the OMIM data.

PREAL: prediction of allergenic protein by maximum Relevance Minimum Redundancy (mRMR) feature selection

BACKGROUND

Assessment of potential allergenicity of protein is necessary whenever transgenic proteins are introduced into the food chain. Bioinformatics approaches in allergen prediction have evolved appreciably in recent years to increase sophistication and performance. However, what are the critical features for protein's allergenicity have been not fully investigated yet.

RESULTS

We presented a more comprehensive model in 128 features space for allergenic proteins prediction by integrating various properties of proteins, such as biochemical and physicochemical properties, sequential features and subcellular locations. The overall accuracy in the cross-validation reached 93.42% to 100% with our new method. Maximum Relevance Minimum Redundancy (mRMR) method and Incremental Feature Selection (IFS) procedure were applied to obtain which features are essential for allergenicity. Results of the performance comparisons showed the superior of our method to the existing methods used widely. More importantly, it was observed that the features of subcellular locations and amino acid composition played major roles in determining the allergenicity of proteins, particularly extracellular/cell surface and vacuole of the subcellular locations for wheat and soybean. To facilitate the allergen prediction, we implemented our computational method in a web application, which can be available at http://gmobl.sjtu.edu.cn/PREAL/index.php.

CONCLUSIONS

Our new approach could improve the accuracy of allergen prediction. And the findings may provide novel insights for the mechanism of allergies.

Prediction of B-cell epitopes using evolutionary information and propensity scales

Background

Development of computational tools that can accurately predict presence and location of B-cell epitopes on pathogenic proteins has a valuable application to the field of vaccinology. Because of the highly variable yet enigmatic nature of B-cell epitopes, their prediction presents a great challenge to computational immunologists.

Methods

We propose a method, BEEPro (B-cell epitope prediction by evolutionary information and propensity scales), which adapts a linear averaging scheme on 16 properties using a support vector machine model to predict both linear and conformational B-cell epitopes. These 16 properties include position specific scoring matrix (PSSM), an amino acid ratio scale, and a set of 14 physicochemical scales obtained via a feature selection process. Finally, a three-way data split procedure is used during the validation process to prevent over-estimation of prediction performance and avoid bias in our experiment results.

Results

In our experiment, first we use a non-redundant linear B-cell epitope dataset curated by Sollner et al. for feature selection and parameter optimization. Evaluated by a three-way data split procedure, BEEPro achieves significant improvement with the area under the receiver operating curve (AUC) = 0.9987, accuracy = 99.29%, mathew's correlation coefficient (MCC) = 0.9281, sensitivity = 0.9604, specificity = 0.9946, positive predictive value (PPV) = 0.9042 for the Sollner dataset. In addition, the same parameters are used to evaluate performance on other independent linear B-cell epitope test datasets, BEEPro attains an AUC which ranges from 0.9874 to 0.9950 and an accuracy which ranges from 93.73% to 97.31%. Moreover, five-fold cross-validation on one benchmark conformational B-cell epitope dataset yields an accuracy of 92.14% and AUC of 0.9066.

Conclusions

Compared with other current models, our method achieves a significant improvement with respect to AUC, accuracy, MCC, sensitivity, specificity, and PPV. Thus, we have shown that an appropriate combination of evolutionary information and propensity scales with a support vector machine model can significantly enhance the prediction performance of both linear and conformational B-cell epitopes.

A novel method based on new adaptive LVQ neural network for predicting protein-protein interactions from protein sequences

Protein-Protein interaction (PPI) is one of the most important data in understanding the cellular processes. Many interesting methods have been proposed in order to predict PPIs. However, the methods which are based on the sequence of proteins as a prior knowledge are more universal. In this paper, a sequence-based, fast, and adaptive PPI prediction method is introduced to assign two proteins to an interaction class (yes, no). First, in order to improve the presentation of the sequences, twelve physicochemical properties of amino acid have been used by different representation methods to transform the sequence of protein pairs into different feature vectors. Then, for speeding up the learning process and reducing the effect of noise PPI data, principal component analysis (PCA) is carried out as a proper feature extraction algorithm. Finally, a new and adaptive Learning Vector Quantization (LVQ) predictor is designed to deal with different models of datasets that are classified into balanced and imbalanced datasets. The accuracy of 93.88%, 90.03%, and 89.72% has been found on S. cerevisiae, H. pylori, and independent datasets, respectively. The results of various experiments indicate the efficiency and validity of the method.

A strategy to select suitable physicochemical attributes of amino acids for protein fold recognition

BACKGROUND

Assigning a protein into one of its folds is a transitional step for discovering three dimensional protein structure, which is a challenging task in bimolecular (biological) science. The present research focuses on: 1) the development of classifiers, and 2) the development of feature extraction techniques based on syntactic and/or physicochemical properties.

RESULTS

Apart from the above two main categories of research, we have shown that the selection of physicochemical attributes of the amino acids is an important step in protein fold recognition and has not been explored adequately. We have presented a multi-dimensional successive feature selection (MD-SFS) approach to systematically select attributes. The proposed method is applied on protein sequence data and an improvement of around 24% in fold recognition has been noted when selecting attributes appropriately.

CONCLUSION

The MD-SFS has been applied successfully in selecting physicochemical attributes of the amino acids. The selected attributes show improved protein fold recognition performance.

Identification of a T7 RNA polymerase variant that permits the enzymatic synthesis of fully 2'-O-methyl-modified RNA

T7 RNA polymerase is an important biocatalyst that is used in diverse biotechnological applications such as in vitro transcription or protein expression. The enzyme displays high substrate specificity which is payed by significant limitations regarding incorporation of synthetic nucleotide analogs. Of specific interest is enzymatic synthesis of 2'-O-methyl-modified RNA as these nucleic acids exhibit improved biochemical and pharmacokinetic properties that make them attractive for diagnostic and therapeutic purposes. We report here on the development of an activity-based selection/screening approach for assessing polymerase activities in the presence of 2'-O-methyl-modified nucleotides, and on the identification of one variant T7 RNA polymerase which is capable of synthesizing all-2'-O-methyl RNA.

Discrimination of soluble and aggregation-prone proteins based on sequence information

Understanding the factors governing protein solubility is a key to grasp the mechanisms of protein solubility and may provide insight into protein aggregation and misfolding related diseases such as Alzheimer's disease. In this work, we attempt to identify factors important to protein solubility using feature selection. Firstly, we calculate 1438 features including physicochemical properties and statistics for each protein. Random Forest algorithm is used to select the most informative and the minimal subset of features based on their predictive performance. A predictive model is built based on 17 selected features. Compared with previous models, our model achieves better performance with a sensitivity of 0.82, specificity 0.85, ACC 0.84, AUC 0.91 and MCC 0.67. Furthermore, a model using a redundancy-reduced dataset (sequence identity <= 30%) achieves the same performance as the model without redundancy reduction. Our results provide not only a reliable model for predicting protein solubility but also a list of features important to protein solubility. The predictive model is implemented as a freely available web application at .

Role of aromatic interactions in amyloid formation by islet amyloid polypeptide

Aromatic-aromatic and aromatic-hydrophobic interactions have been proposed to play a role in amyloid formation by a range of polypeptides, including islet amyloid polypeptide (IAPP or amylin). IAPP is responsible for amyloid formation in patients with type 2 diabetes. The polypeptide is 37 residues long and contains three aromatic residues, Phe-15, Phe-23, and Tyr-37. The ability of all single aromatic to leucine mutants, all double aromatic to leucine mutants, and the triple leucine mutant to form amyloid were examined. Amyloid formation was almost twice as rapid for the F15L mutant as for the wild type but was almost 3-fold slower for the Y37L mutant and almost 2-fold slower for the F23L mutant. Amyloid fibrils formed from each of the single mutants were effective at seeding amyloid formation by wild-type IAPP, implying that the fibril structures are similar. The F15L/F23L double mutant has a larger effect than the F15L/Y37L double mutant on the rate of amyloid formation, even though a Y37L substitution has more drastic consequences in the wild-type background than does the F23L mutation, suggesting nonadditive effects between the different sites. The triple leucine mutant and the F23L/Y37L double mutant are the slowest to form amyloid. F15 has been proposed to make important contacts early in the aggregation pathway, but the data for the F15L mutant indicate that they are not optimal. A set of variants containing natural and unnatural amino acids at position 15, which were designed to conserve hydrophobicity, but alter α-helix and β-sheet propensity, were analyzed to determine the properties of this position that control the rate of amyloid formation. There is no correlation between β-sheet propensity at this position and the rate of amyloid formation, but there is a correlation with α-helical propensity.

Construction of protein distance matrix based on amino acid indices and Discrete Fourier Transform

Protein distance matrix is widely used in various protein sequence analyses, and mainly obtained by using pairwise sequence alignment scores or protein sequence homology, which fail to take into consideration of individual physical characteristics of protein sequences and amino acids, or a combination of these features. In this paper, a new method is therefore proposed for constructing protein distance matrix based on natural amino acid indices in combination with Discrete Fourier Transform (DFT). For the proposed method, protein distance matrices can be generated using any given set of amino acid indices, each one of which represents a unique biological feature of protein sequences. In this study, the results are based on the combination of 25 widely accepted amino acid indices, which produced the best results, according to the biological relationships between proteins. As a case study 26 Cluster of Differentiation 4 (CD4) protein sequences were used in order to construct a distance matrix based on the proposed method. The results show that the pairwise relationship between CD4 protein sequences remain the same in comparison with their pairwise percent identity. For another group of protein sequences the pairwise relationship between CD4 protein sequences dramatically changed with the proposed method in comparison to the pairwise percent identity. The proposed distance matrix has been shown to have a positive impact on these case studies and therefore is expected to be useful in several fields such as multiple protein sequence alignment and phylogenetic analysis, where an accurate distance matrix based on natural generalized protein properties plays an important role.

Self-assembly of amphipathic β-sheet peptides: insights and applications

Amphipathic peptides composed of alternating polar and nonpolar residues have a strong tendency to self-assemble into one-dimensional, amyloid-like fibril structures. Fibrils derived from peptides of general (XZXZ)(n) sequence in which X is hydrophobic and Z is hydrophilic adopt a putative β-sheet bilayer. The bilayer configuration allows burial of the hydrophobic X side chain groups in the core of the fibril and leaves the polar Z side chains exposed to solvent. This architectural arrangement provides fibrils that maintain high solubility in water and has facilitated the recent exploitation of self-assembled amphipathic peptide fibrils as functional biomaterials. This article is a critical review of the development and application of self-assembling amphipathic peptides with a focus on the fundamental insight these types of peptides provide into peptide self-assembly phenomena.

PMeS: prediction of methylation sites based on enhanced feature encoding scheme

Protein methylation is predominantly found on lysine and arginine residues, and carries many important biological functions, including gene regulation and signal transduction. Given their important involvement in gene expression, protein methylation and their regulatory enzymes are implicated in a variety of human disease states such as cancer, coronary heart disease and neurodegenerative disorders. Thus, identification of methylation sites can be very helpful for the drug designs of various related diseases. In this study, we developed a method called PMeS to improve the prediction of protein methylation sites based on an enhanced feature encoding scheme and support vector machine. The enhanced feature encoding scheme was composed of the sparse property coding, normalized van der Waals volume, position weight amino acid composition and accessible surface area. The PMeS achieved a promising performance with a sensitivity of 92.45%, a specificity of 93.18%, an accuracy of 92.82% and a Matthew’s correlation coefficient of 85.69% for arginine as well as a sensitivity of 84.38%, a specificity of 93.94%, an accuracy of 89.16% and a Matthew’s correlation coefficient of 78.68% for lysine in 10-fold cross validation. Compared with other existing methods, the PMeS provides better predictive performance and greater robustness. It can be anticipated that the PMeS might be useful to guide future experiments needed to identify potential methylation sites in proteins of interest. The online service is available at http://bioinfo.ncu.edu.cn/inquiries_PMeS.aspx.

PDZ domain-containing 1 (PDZK1) protein regulates phospholipase C-β3 (PLC-β3)-specific activation of somatostatin by forming a ternary complex with PLC-β3 and somatostatin receptors

Phospholipase C-β (PLC-β) is a key molecule in G protein-coupled receptor (GPCR)-mediated signaling. Many studies have shown that the four PLC-β subtypes have different physiological functions despite their similar structures. Because the PLC-β subtypes possess different PDZ-binding motifs, they have the potential to interact with different PDZ proteins. In this study, we identified PDZ domain-containing 1 (PDZK1) as a PDZ protein that specifically interacts with PLC-β3. To elucidate the functional roles of PDZK1, we next screened for potential interacting proteins of PDZK1 and identified the somatostatin receptors (SSTRs) as another protein that interacts with PDZK1. Through these interactions, PDZK1 assembles as a ternary complex with PLC-β3 and SSTRs. Interestingly, the expression of PDZK1 and PLC-β3, but not PLC-β1, markedly potentiated SST-induced PLC activation. However, disruption of the ternary complex inhibited SST-induced PLC activation, which suggests that PDZK1-mediated complex formation is required for the specific activation of PLC-β3 by SST. Consistent with this observation, the knockdown of PDZK1 or PLC-β3, but not that of PLC-β1, significantly inhibited SST-induced intracellular Ca(2+) mobilization, which further attenuated subsequent ERK1/2 phosphorylation. Taken together, our results strongly suggest that the formation of a complex between SSTRs, PDZK1, and PLC-β3 is essential for the specific activation of PLC-β3 and the subsequent physiologic responses by SST.

Computational prediction of furin cleavage sites by a hybrid method and understanding mechanism underlying diseases

Furin cleaves diverse types of protein precursors in the secretory pathway. The substrates for furin cleavage possess a specific 20-residue recognition sequence motif. In this report, based on the functional characterisation of the 20-residue sequence motif, we developed a furin cleavage site prediction tool, PiTou, using a hybrid method composed of a hidden Markov model and biological knowledge-based cumulative probability score functions. PiTou can accurately predict the presence and location of furin cleavage sites in protein sequences with high sensitivity (96.9%) and high specificity (97.3%). PiTou's prediction scores are biological meaningful and reflect binding strength and solvent accessibility of furin substrates. A prediction result is interpreted within cellular contexts: subcellular localisation, cellular function and interference by other dynamic protein modifications. Combining next-generation sequencing, PiTou can help with elucidating the molecular mechanism of furin cleavage-associated human diseases. PiTou has been made freely available at the associated website.

Rewiring of PDZ domain-ligand interaction network contributed to eukaryotic evolution

PDZ domain-mediated interactions have greatly expanded during metazoan evolution, becoming important for controlling signal flow via the assembly of multiple signaling components. The evolutionary history of PDZ domain-mediated interactions has never been explored at the molecular level. It is of great interest to understand how PDZ domain-ligand interactions emerged and how they become rewired during evolution. Here, we constructed the first human PDZ domain-ligand interaction network (PDZNet) together with binding motif sequences and interaction strengths of ligands. PDZNet includes 1,213 interactions between 97 human PDZ proteins and 591 ligands that connect most PDZ protein-mediated interactions (98%) in a large single network via shared ligands. We examined the rewiring of PDZ domain-ligand interactions throughout eukaryotic evolution by tracing changes in the C-terminal binding motif sequences of the PDZ ligands. We found that interaction rewiring by sequence mutation frequently occurred throughout evolution, largely contributing to the growth of PDZNet. The rewiring of PDZ domain-ligand interactions provided an effective means of functional innovations in nervous system development. Our findings provide empirical evidence for a network evolution model that highlights the rewiring of interactions as a mechanism for the development of new protein functions. PDZNet will be a valuable resource to further characterize the organization of the PDZ domain-mediated signaling proteome.

Rewiring of interactions is a powerful tool for the evolution of organism complexity. Rewiring among preexisting proteins provides a simple mechanism for the development of new signaling circuits by redirecting information flows without a gain or loss of genes. Particularly, interactions mediated by short linear motifs can be easily changed by mutations during evolution, resulting in a rewiring of interactions. However, how interaction rewiring of linear motif interactions facilitates the emergence of new protein function during evolution is poorly understood. Here, we systematically investigated the rewiring of interactions mediated by PDZ domains, which are one of the most commonly found peptide recognition modules. We found that PDZ domain-ligand interactions are frequently rewired by C-terminal sequence mutations in PDZ ligands during evolution. Especially, rewiring of PDZ domain-ligand interactions was involved in neuronal function development, occurring concurrently with the emergence of vertebrates and suggesting that reorganization of signaling pathways by rewiring PDZ domain-ligand interactions significantly contributed to the evolution of nervous systems in vertebrates. Our findings highlight the rewiring of interactions as an effective means for functional innovation, providing new insight into eukaryotic evolution, which has not been fully explained by only the expansion of protein families.

Self-assembled amino acids and dipeptides as noncovalent hydrogels for tissue engineering

This review critically assesses progress in the use of self-assembling dipeptides and amino acids as hydrogel materials for tissue engineering.

Label-free protein quantitation using weighted spectral counting

Mass spectrometry (MS)-based shotgun proteomics allows protein identifications even in complex biological samples. Protein abundances can then be estimated from the counts of MS/MS spectra attributable to each protein, provided that one corrects for differential MS-detectability of the contributing peptides. We describe the use of a method, APEX, which calculates Absolute Protein EXpression levels based on learned correction factors, MS/MS spectral counts, and each protein's probability of correct identification.The APEX-based calculations consist of three parts: (1) Using training data, peptide sequences and their sequence properties, a model is built that can be used to estimate MS-detectability (O (i)) for any given protein. (2) Absolute abundances of proteins measured in an MS/MS experiment are calculated with information from spectral counts, identification probabilities and the learned O (i)-values. (3) Simple statistics allow for significance analysis of differential expression in two distinct biological samples, i.e., measuring relative protein abundances. APEX-based protein abundances span more than four orders of magnitude and are applicable to mixtures of hundreds to thousands of proteins from any type of organism.

POPISK: T-cell reactivity prediction using support vector machines and string kernels

Background

Accurate prediction of peptide immunogenicity and characterization of relation between peptide sequences and peptide immunogenicity will be greatly helpful for vaccine designs and understanding of the immune system. In contrast to the prediction of antigen processing and presentation pathway, the prediction of subsequent T-cell reactivity is a much harder topic. Previous studies of identifying T-cell receptor (TCR) recognition positions were based on small-scale analyses using only a few peptides and concluded different recognition positions such as positions 4, 6 and 8 of peptides with length 9. Large-scale analyses are necessary to better characterize the effect of peptide sequence variations on T-cell reactivity and design predictors of a peptide's T-cell reactivity (and thus immunogenicity). The identification and characterization of important positions influencing T-cell reactivity will provide insights into the underlying mechanism of immunogenicity.

Results

This work establishes a large dataset by collecting immunogenicity data from three major immunology databases. In order to consider the effect of MHC restriction, peptides are classified by their associated MHC alleles. Subsequently, a computational method (named POPISK) using support vector machine with a weighted degree string kernel is proposed to predict T-cell reactivity and identify important recognition positions. POPISK yields a mean 10-fold cross-validation accuracy of 68% in predicting T-cell reactivity of HLA-A2-binding peptides. POPISK is capable of predicting immunogenicity with scores that can also correctly predict the change in T-cell reactivity related to point mutations in epitopes reported in previous studies using crystal structures. Thorough analyses of the prediction results identify the important positions 4, 6, 8 and 9, and yield insights into the molecular basis for TCR recognition. Finally, we relate this finding to physicochemical properties and structural features of the MHC-peptide-TCR interaction.

Conclusions

A computational method POPISK is proposed to predict immunogenicity with scores which are useful for predicting immunogenicity changes made by single-residue modifications. The web server of POPISK is freely available at http://iclab.life.nctu.edu.tw/POPISK.

Structural requirements for interaction of peroxisomal targeting signal 2 and its receptor PEX7

The import of a subset of peroxisomal matrix proteins is mediated by the peroxisomal targeting signal 2 (PTS2). The results of our sequence and physical property analysis of known PTS2 signals and of a mutational study of the least characterized amino acids of a canonical PTS2 motif indicate that PTS2 forms an amphipathic helix accumulating all conserved residues on one side. Three-dimensional structural modeling of the PTS2 receptor PEX7 reveals a groove with an evolutionarily conserved charge distribution complementary to PTS2 signals. Mammalian two-hybrid assays and cross-complementation of a mutation in PTS2 by a compensatory mutation in PEX7 confirm the interaction site. An unstructured linker region separates the PTS2 signal from the core protein. This additional information on PTS2 signals was used to generate a PTS2 prediction algorithm that enabled us to identify novel PTS2 signals within human proteins and to describe KChIP4 as a novel peroxisomal protein.

Complementary π-π interactions induce multicomponent coassembly into functional fibrils

Noncovalent self-assembled materials inspired by amyloid architectures are useful for biomedical applications ranging from regenerative medicine to drug delivery. The selective coassembly of complementary monomeric units to provide ordered multicomponent fibrils is a possible strategy for enhancing the sophistication of these noncovalent materials. Herein we report that complementary π-π interactions can be exploited to promote the coassembly of phenylalanine (Phe) derivatives that possess complementary aromatic side-chain functionality. Specifically, equimolar mixtures of Fmoc-Phe and Fmoc-F(5)-Phe, which possess side-chain groups with complementary quadrupole electronics, readily coassemble to form two-component fibrils and hydrogels under conditions where Fmoc-Phe alone fails to self-assemble. In addition, it was found that equimolar mixtures of Fmoc-Phe with monohalogenated (F, Cl, and Br) Fmoc-Phe derivatives also coassembled into two-component fibrils. These results collectively indicate that face-to-face quadrupole stacking between benzyl side-chain groups does not account for the molecular recognition between Phe and halogenated Phe derivatives that promote cofibrillization but that coassembly is mediated by more subtle π-π effects arising from the halogenation of the benzyl side chain. The use of complementary π-π interactions to promote the coassembly of two distinct monomeric units into ordered two-component fibrils dramatically expands the repertoire of noncovalent interactions that can be used in the development of sophisticated noncovalent materials.

Epidemiology and mutational analysis of global strains of Crimean-Congo haemorrhagic fever virus

Crimean-Congo hemorrhagic fever (CCHF) is a severe illness with high fatality. Cases are reported in several countries in Africa, Europe, the Middle East, and Asia. Phylogenetic analyses based on the virus S (nucleocapsid), M (glycoprotein), and L (polymerase) genome segments sequences indicate distinct geographic lineages exist but their specific genetic characteristics require elucidation. In this work we collected all full length S segment sequences and generated a phylogenetic tree based on the alignment of these 62 samples. We then analyzed the alignment using entries from AAIndex, the Amino Acid Index database, to identify amino acid mutations that performed significant changes in charge, pka, hydropathy and side chain volume. Finally, we mapped these changes back to the tree and alignment to identify correlated mutations or sites that characterized a specific lineage. Based on this analysis we are able to propose a number of sites that appear to be important for virus function and which would be good candidates for experimental mutational analysis studies.

Rapid probing of biological surfaces with a sparse-matrix peptide library

Finding unique peptides to target specific biological surfaces is crucial to basic research and technology development, though methods based on biological arrays or large libraries limit the speed and ease with which these necessary compounds can be found. We reasoned that because biological surfaces, such as cell surfaces, mineralized tissues, and various extracellular matrices have unique molecular compositions, they present unique physicochemical signatures to the surrounding medium which could be probed by peptides with appropriately corresponding physicochemical properties. To test this hypothesis, a naïve pilot library of 36 peptides, varying in their hydrophobicity and charge, was arranged in a two-dimensional matrix and screened against various biological surfaces. While the number of peptides in the matrix library was very small, we obtained “hits” against all biological surfaces probed. Sequence refinement of the “hits” led to peptides with markedly higher specificity and binding activity against screened biological surfaces. Genetic studies revealed that peptide binding to bacteria was mediated, at least in some cases, by specific cell-surface molecules, while examination of human tooth sections showed that this method can be used to derive peptides with highly specific binding to human tissue.

SNOSite: exploiting maximal dependence decomposition to identify cysteine S-nitrosylation with substrate site specificity

S-nitrosylation, the covalent attachment of a nitric oxide to (NO) the sulfur atom of cysteine, is a selective and reversible protein post-translational modification (PTM) that regulates protein activity, localization, and stability. Despite its implication in the regulation of protein functions and cell signaling, the substrate specificity of cysteine S-nitrosylation remains unknown. Based on a total of 586 experimentally identified S-nitrosylation sites from SNAP/L-cysteine-stimulated mouse endothelial cells, this work presents an informatics investigation on S-nitrosylation sites including structural factors such as the flanking amino acids composition, the accessible surface area (ASA) and physicochemical properties, i.e. positive charge and side chain interaction parameter. Due to the difficulty to obtain the conserved motifs by conventional motif analysis, maximal dependence decomposition (MDD) has been applied to obtain statistically significant conserved motifs. Support vector machine (SVM) is applied to generate predictive model for each MDD-clustered motif. According to five-fold cross-validation, the MDD-clustered SVMs could achieve an accuracy of 0.902, and provides a promising performance in an independent test set. The effectiveness of the model was demonstrated on the correct identification of previously reported S-nitrosylation sites of Bos taurus dimethylarginine dimethylaminohydrolase 1 (DDAH1) and human hemoglobin subunit beta (HBB). Finally, the MDD-clustered model was adopted to construct an effective web-based tool, named SNOSite (http://csb.cse.yzu.edu.tw/SNOSite/), for identifying S-nitrosylation sites on the uncharacterized protein sequences.

Prediction of the exposure status of transmembrane beta barrel residues from protein sequence

We present BTMX (Beta barrel TransMembrane eXposure), a computational method to predict the exposure status (i.e. exposed to the bilayer or hidden in the protein structure) of transmembrane residues in transmembrane beta barrel proteins (TMBs). BTMX predicts the exposure status of known TM residues with an accuracy of 84.2% over 2,225 residues and provides a confidence score for all predictions. Predictions made are in concert with the fact that hydrophobic residues tend to be more exposed to the bilayer. The biological relevance of the input parameters is also discussed. The highest prediction accuracy is obtained when a sliding window comprising three residues with similar C(α)-C(β) vector orientations is employed. The prediction accuracy of the BTMX method on a separate unseen non-redundant test dataset is 78.1%. By employing out-pointing residues that are exposed to the bilayer, we have identified various physico-chemical properties that show statistically significant differences between the beta strands located at the oligomeric interfaces compared to the non-oligomeric strands. The BTMX web server generates colored, annotated snake-plots as part of the prediction results and is available under the BTMX tab at http://service.bioinformatik.uni-saarland.de/tmx-site/. Exposure status prediction of TMB residues may be useful in 3D structure prediction of TMBs.

Enhancement of HIV-1 infectivity by simple, self-assembling modular peptides

Semen-derived enhancer of viral infection (SEVI), an amyloid fibril formed from a cationic peptide fragment of prostatic acidic phosphatase (PAP), dramatically enhances the infectivity of human immunodeficiency virus type 1 (HIV-1). Insoluble, sedimentable fibrils contribute to SEVI-mediated enhancement of virus infection. However, the SEVI-forming PAP(248-286) peptide is able to produce infection-enhancing structures much more quickly than it forms amyloid fibrils. This suggests that soluble supramolecular assemblies may enhance HIV-1 infection. To address this question, non-SEVI amyloid-like fibrils were derived from general amphipathic peptides of sequence Ac-K(n)(XKXE)(2)-NH(2). These cationic peptides efficiently self-assembled to form soluble, fibril-like structures that were, in some cases, able to enhance HIV-1 infection even more efficiently than SEVI. Experiments were also performed to determine whether agents that efficiently shield the charged surface of SEVI fibrils block SEVI-mediated infection-enhancement. To do this, we generated self-assembling anionic peptides of sequence Ac-E(n)(XKXE)(2)-NH(2). One of these peptides completely abrogated SEVI-mediated enhancement of HIV-1 infection, without altering HIV-1 infectivity in the absence of SEVI. Collectively, these data suggest that soluble SEVI assemblies may mediate infection-enhancement, and that anionic peptide supramolecular assemblies have the potential to act as anti-SEVI microbicides.

Tuning β-sheet peptide self-assembly and hydrogelation behavior by modification of sequence hydrophobicity and aromaticity

Peptide self-assembly leading to cross-β amyloid structures is a widely studied phenomenon because of its role in amyloid pathology and the exploitation of amyloid as a functional biomaterial. The self-assembly process is governed by hydrogen bonding, hydrophobic, aromatic π-π, and electrostatic Coulombic interactions. A role for aromatic π-π interactions in peptide self-assembly leading to amyloid has been proposed, but the relative contributions of π-π versus general hydrophobic interactions in these processes are poorly understood. The Ac-(XKXK)(2)-NH(2) peptide was used to study the contributions of aromatic and hydrophobic interactions to peptide self-assembly. Position X was globally replaced by valine (Val), isoleucine (Ile), phenylalanine (Phe), pentafluorophenylalanine (F(5)-Phe), and cyclohexylalanine (Cha). At low pH, these peptides remain monomeric because of repulsion of charged lysine (Lys) residues. Increasing the solvent ionic strength to shield repulsive charge-charge interactions between protonated Lys residues facilitated cross-β fibril formation. It was generally found that as peptide hydrophobicity increased, the required ionic strength to induce self-assembly decreased. At [NaCl] ranging from 0 to 1000 mM, the Val sequence failed to assemble. Assembly of the Phe sequence commenced at 700 mM NaCl and at 300 mM NaCl for the less hydrophobic Ile variant, even though it displayed a mixture of random coil and β-sheet secondary structures over all NaCl concentrations. β-Sheet formation for F(5)-Phe and Cha sequences was observed at only 20 and 60 mM NaCl, respectively. Whereas self-assembly propensity generally correlated to peptide hydrophobicity and not aromatic character the presence of aromatic amino acids imparted unique properties to fibrils derived from these peptides. Nonaromatic peptides formed fibrils of 3-15 nm in diameter, whereas aromatic peptides formed nanotape or nanoribbon architectures of 3-7 nm widths. In addition, all peptides formed fibrillar hydrogels at sufficient peptide concentrations, but nonaromatic peptides formed weak gels, whereas aromatic peptides formed rigid gels. These findings clarify the influence of aromatic amino acids on peptide self-assembly processes and illuminate design principles for the inclusion of aromatic amino acids in amyloid-derived biomaterials.

Statistical analysis and exposure status classification of transmembrane beta barrel residues

Several computational methods exist for the identification of transmembrane beta barrel proteins (TMBs) from sequence. Some of these methods also provide the transmembrane (TM) boundaries of the putative TMBs. The aim of this study is to (1) derive the propensities of the TM residues to be exposed to the lipid bilayer and (2) to predict the exposure status (i.e. exposed to the bilayer or hidden in protein structure) of TMB residues. Three novel propensity scales namely, BTMC, BTMI and HTMI were derived for the TMB residues at the hydrophobic core region of the outer membrane (OM), the lipid-water interface regions of the OM, and for the helical membrane proteins (HMPs) residues at the lipid-water interface regions of the inner membrane (IM), respectively. Separate propensity scales were derived for monomeric and functionally oligomeric TMBs. The derived propensities reflect differing physico-chemical properties of the respective membrane bilayer regions and were employed in a computational method for the prediction of the exposure status of TMB residues. Based on the these propensities, the conservation indices and the frequency profile of the residues, the transmembrane residues were classified into buried/exposed with an accuracy of 77.91% and 80.42% for the residues at the membrane core and the interface regions, respectively. The correlation of the derived scales with different physico-chemical properties obtained from the AAIndex database are also discussed. Knowledge about the residue propensities and burial status will be useful in annotating putative TMBs with unknown structure.

Incorporating distant sequence features and radial basis function networks to identify ubiquitin conjugation sites

Ubiquitin (Ub) is a small protein that consists of 76 amino acids about 8.5 kDa. In ubiquitin conjugation, the ubiquitin is majorly conjugated on the lysine residue of protein by Ub-ligating (E3) enzymes. Three major enzymes participate in ubiquitin conjugation. They are E1, E2 and E3 which are responsible for activating, conjugating and ligating ubiquitin, respectively. Ubiquitin conjugation in eukaryotes is an important mechanism of the proteasome-mediated degradation of a protein and regulating the activity of transcription factors. Motivated by the importance of ubiquitin conjugation in biological processes, this investigation develops a method, UbSite, which uses utilizes an efficient radial basis function (RBF) network to identify protein ubiquitin conjugation (ubiquitylation) sites. This work not only investigates the amino acid composition but also the structural characteristics, physicochemical properties, and evolutionary information of amino acids around ubiquitylation (Ub) sites. With reference to the pathway of ubiquitin conjugation, the substrate sites for E3 recognition, which are distant from ubiquitylation sites, are investigated. The measurement of F-score in a large window size (-20∼+20) revealed a statistically significant amino acid composition and position-specific scoring matrix (evolutionary information), which are mainly located distant from Ub sites. The distant information can be used effectively to differentiate Ub sites from non-Ub sites. As determined by five-fold cross-validation, the model that was trained using the combination of amino acid composition and evolutionary information performs best in identifying ubiquitin conjugation sites. The prediction sensitivity, specificity, and accuracy are 65.5%, 74.8%, and 74.5%, respectively. Although the amino acid sequences around the ubiquitin conjugation sites do not contain conserved motifs, the cross-validation result indicates that the integration of distant sequence features of Ub sites can improve predictive performance. Additionally, the independent test demonstrates that the proposed method can outperform other ubiquitylation prediction tools.

Gain of local structure in an amphipathic peptide does not require a specific tertiary framework

In this work, we studied how an amphipathic peptide of the surface of the globular protein thioredoxin, TRX94-108, acquires a native-like structure when it becomes involved in an apolar interaction network. We designed peptide variants where the tendency to form alpha-helical conformation is modulated by replacing each of the leucine amino acid residues by an alanine. The induction of structure caused by sodium dodecyl sulfate (SDS) binding was studied by capillary zone electrophoresis, circular dichroism, DOSY-NMR, and molecular dynamics simulations (MDS). In addition, we analyzed the strength of the interaction between a C18 RP-HPLC matrix and the peptides. The results presented here reveal that (a) critical elements in the sequence of the wild-type peptide stabilize a SDS/peptide supramolecular cluster; (b) the hydrophobic nature of the interaction between SDS molecules and the peptide constrains the ensemble of conformations; (c) nonspecific apolar surfaces are sufficient to stabilize peptide secondary structure. Remarkably, MDS shed light on a contact network formed by a limited number of SDS molecules that serves as a structural scaffold preserving the helical conformation of this module. This mechanism might prevail when a peptide with low helical propensity is involved in structure consolidation. We suggest that folding of peptides sharing this feature does not require a preformed tightly-packed protein core. Thus, the formation of specific tertiary interactions would be the consequence of peptide folding and not its cause. In this scenario, folding might be thought of as a process that includes unspecific rounds of structure stabilization guiding the protein to the native state.

Amino acid sequence preferences to control cell-specific organization of endothelial cells, smooth muscle cells, and fibroblasts

Effective surface modification with biocompatible molecules is known to be effective in reducing the life-threatening risks related to artificial cardiovascular implants. In recent strategies in regenerative medicine, the enhancement and support of natural repair systems at the site of injury by designed biocompatible molecules have succeeded in rapid and effective injury repair. Therefore, such a strategy could also be effective for rapid endothelialization of cardiovascular implants to lower the risk of thrombosis and stenosis. To achieve this enhancement of the natural repair system, a biomimetic molecule that mimics proper cellular organization at the implant location is required. In spite of the fact that many reported peptides have cell-attracting properties on material surfaces, there have been few peptides that could control cell-specific adhesion. For the advanced cardiovascular implants, peptides that can mimic the natural mechanism that controls cell-specific organization have been strongly anticipated. To obtain such peptides, we hypothesized the cellular bias toward certain varieties of amino acids and examined the cell preference (in terms of adhesion, proliferation, and protein attraction) of varieties and of repeat length on SPOT peptide arrays. To investigate the role of specific peptides in controlling the organization of various cardiovascular-related cells, we compared endothelial cells (ECs), smooth muscle cells (SMCs), and fibroblasts (FBs). A clear, cell-specific preference was found for amino acids (longer than 5-mer) using three types of cells, and the combinational effect of the physicochemical properties of the residues was analyzed to interpret the mechanism.

The substrate specificity profile of human granzyme A

The exact biological function of granzyme A, a granule-associated serine protease belonging to the tryptase family of proteases, is still a matter of debate because conflicting roles have been suggested, such as initiation of caspase-independent apoptosis-like cell death and endogenous modulation of inflammatory processes. In contrast to its well-studied family member, granzyme B, far less is known about the physiological targets of granzyme A. Using an N-terminal peptide-centric proteomics technology, the substrate specificity of human granzyme A was extensively characterized at the level of macromolecular protein substrates. Overall, more than 260 cleavage sites, almost exclusively favoring basic residues at the P1 position, in approximately 200 unique protein substrates, including the well-known in vitro substrates APEX-endonuclease 1 and different histones, were identified. Further substrate characterization was used to delineate physical properties in the substrate specificity profiles, which further highlights important aspects in protease/substrate biology.

Analysis and prediction of translation rate based on sequence and functional features of the mRNA

Protein concentrations depend not only on the mRNA level, but also on the translation rate and the degradation rate. Prediction of mRNA's translation rate would provide valuable information for in-depth understanding of the translation mechanism and dynamic proteome. In this study, we developed a new computational model to predict the translation rate, featured by (1) integrating various sequence-derived and functional features, (2) applying the maximum relevance & minimum redundancy method and incremental feature selection to select features to optimize the prediction model, and (3) being able to predict the translation rate of RNA into high or low translation rate category. The prediction accuracies under rich and starvation condition were 68.8% and 70.0%, respectively, evaluated by jackknife cross-validation. It was found that the following features were correlated with translation rate: codon usage frequency, some gene ontology enrichment scores, number of RNA binding proteins known to bind its mRNA product, coding sequence length, protein abundance and 5′UTR free energy. These findings might provide useful information for understanding the mechanisms of translation and dynamic proteome. Our translation rate prediction model might become a high throughput tool for annotating the translation rate of mRNAs in large-scale.