Proteins with complex architecture as potential targets for drug design: a case study of Mycobacterium tuberculosis

Lengthy co-evolution of Homo sapiens and Mycobacterium tuberculosis, the main causative agent of tuberculosis, resulted in a dramatically successful pathogen species that presents considerable challenge for modern medicine. The continuous and ever increasing appearance of multi-drug resistant mycobacteria necessitates the identification of novel drug targets and drugs with new mechanisms of action. However, further insights are needed to establish automated protocols for target selection based on the available complete genome sequences. In the present study, we perform complete proteome level comparisons between M. tuberculosis, mycobacteria, other prokaryotes and available eukaryotes based on protein domains, local sequence similarities and protein disorder. We show that the enrichment of certain domains in the genome can indicate an important function specific to M. tuberculosis. We identified two families, termed pkn and PE/PPE that stand out in this respect. The common property of these two protein families is a complex domain organization that combines species-specific regions, commonly occurring domains and disordered segments. Besides highlighting promising novel drug target candidates in M. tuberculosis, the presented analysis can also be viewed as a general protocol to identify proteins involved in species-specific functions in a given organism. We conclude that target selection protocols should be extended to include proteins with complex domain architectures instead of focusing on sequentially unique and essential proteins only.     

Identification of protein domains by shotgun proteolysis

The identification of protein domains within multi-domain proteins is a persistent problem. Here, we describe an experimental method (shotgun proteolysis) based on random DNA fragmentation and protease selection of the encoded polypeptides on phage for this purpose. We applied the method to the Escherichia coli genome and identified 124 protease-resistant fragments; several were re-cloned for expression as soluble fragments in bacteria, and corresponded to autonomously folding units with folding energies similar to natural protein domains (DeltaG(u)=3.8-6.6 kcal/mol). Structural information was available for approximately half of the selected proteins, which corresponded to compact, globular and domain-sized units that had been derived from a wide range of protein superfamilies. Furthermore, boundaries of the selected fragments correlated with domain boundaries as defined by bioinformatics predictions (R2=0.82; p=0.016). However, predictions were incomplete or entirely lacking for the remaining fragments, reflecting the limited proteome coverage of current bioinformatics methods. Shotgun proteolysis therefore provides a means to identify domains and other autonomously folding units on a genome-wide scale, without any prior knowledge of sequence or structure. Shotgun proteolysis should be particularly valuable for structural studies of proteins and represents a high-throughput alternative to the classical limited proteolysis method for the isolation of stable components of multi-domain proteins.     

FIST: a sensory domain for diverse signal transduction pathways in prokaryotes and ubiquitin signaling in eukaryotes

MOTIVATION: Sensory domains that are conserved among Bacteria, Archaea and Eucarya are important detectors of common signals detected by living cells. Due to their high sequence divergence, sensory domains are difficult to identify. We systematically look for novel sensory domains using sensitive profile-based searches initiated with regions of signal transduction proteins where no known domains can be identified by current domain models. RESULTS: Using profile searches followed by multiple sequence alignment, structure prediction and domain architecture analysis, we have identified a novel sensory domain termed FIST, which is present in signal transduction proteins from Bacteria, Archaea and Eucarya. Chromosomal proximity of FIST-encoding genes to those coding for proteins involved in amino acid metabolism and transport suggest that FIST domains bind small ligands, such as amino acids.     

In Silico Characterization of Proteins: UniProt,InterPro and Integr8

Nucleic acid sequences from genome sequencing projects are submitted as raw data, from which biologists attempt to elucidate the function of the predicted gene products. The protein sequences are stored in public databases, such as the UniProt Knowledgebase (UniProtKB), where curators try to add predicted and experimental functional information. Protein function prediction can be done using sequence similarity searches, but an alternative approach is to use protein signatures, which classify proteins into families and domains. The major protein signature databases are available through the integrated InterPro database, which provides a classification of UniProtKB sequences. As well as characterization of proteins through protein families, many researchers are interested in analyzing the complete set of proteins from a genome (i.e. the proteome), and there are databases and resources that provide non-redundant proteome sets and analyses of proteins from organisms with completely sequenced genomes. This article reviews the tools and resources available on the web for single and large-scale protein characterization and whole proteome analysis.     

Uncovering the human methyltransferasome

We present a comprehensive analysis of the human methyltransferasome. Primary sequences, predicted secondary structures, and solved crystal structures of known methyltransferases were analyzed by hidden Markov models, Fisher-based statistical matrices, and fold recognition prediction-based threading algorithms to create a model, or profile, of each methyltransferase superfamily. These profiles were used to scan the human proteome database and detect novel methyltransferases. 208 proteins in the human genome are now identified as known or putative methyltransferases, including 38 proteins that were not annotated previously. To date, 30% of these proteins have been linked to disease states. Possible substrates of methylation for all of the SET domain and SPOUT methyltransferases as well as 100 of the 131 seven-β-strand methyltransferases were surmised from sequence similarity clusters based on alignments of the substrate-specific domains.     

DomPep--a general method for predicting modular domain-mediated protein-protein interactions

Protein-protein interactions (PPIs) are frequently mediated by the binding of a modular domain in one protein to a short, linear peptide motif in its partner. The advent of proteomic methods such as peptide and protein arrays has led to the accumulation of a wealth of interaction data for modular interaction domains. Although several computational programs have been developed to predict modular domain-mediated PPI events, they are often restricted to a given domain type. We describe DomPep, a method that can potentially be used to predict PPIs mediated by any modular domains. DomPep combines proteomic data with sequence information to achieve high accuracy and high coverage in PPI prediction. Proteomic binding data were employed to determine a simple yet novel parameter Ligand-Binding Similarity which, in turn, is used to calibrate Domain Sequence Identity and Position-Weighted-Matrix distance, two parameters that are used in constructing prediction models. Moreover, DomPep can be used to predict PPIs for both domains with experimental binding data and those without. Using the PDZ and SH2 domain families as test cases, we show that DomPep can predict PPIs with accuracies superior to existing methods. To evaluate DomPep as a discovery tool, we deployed DomPep to identify interactions mediated by three human PDZ domains. Subsequent in-solution binding assays validated the high accuracy of DomPep in predicting authentic PPIs at the proteome scale. Because DomPep makes use of only interaction data and the primary sequence of a domain, it can be readily expanded to include other types of modular domains.     

Armadillo: domain boundary prediction by amino acid composition

The identification and annotation of protein domains provides a critical step in the accurate determination of molecular function. Both computational and experimental methods of protein structure determination may be deterred by large multi-domain proteins or flexible linker regions. Knowledge of domains and their boundaries may reduce the experimental cost of protein structure determination by allowing researchers to work on a set of smaller and possibly more successful alternatives. Current domain prediction methods often rely on sequence similarity to conserved domains and as such are poorly suited to detect domain structure in poorly conserved or orphan proteins. We present here a simple computational method to identify protein domain linkers and their boundaries from sequence information alone. Our domain predictor, Armadillo (http://armadillo.blueprint.org), uses any amino acid index to convert a protein sequence to a smoothed numeric profile from which domains and domain boundaries may be predicted. We derived an amino acid index called the domain linker propensity index (DLI) from the amino acid composition of domain linkers using a non-redundant structure dataset. The index indicates that Pro and Gly show a propensity for linker residues while small hydrophobic residues do not. Armadillo predicts domain linker boundaries from Z-score distributions and obtains 35% sensitivity with DLI in a two-domain, single-linker dataset (within +/-20 residues from linker). The combination of DLI and an entropy-based amino acid index increases the overall Armadillo sensitivity to 56% for two domain proteins. Moreover, Armadillo achieves 37% sensitivity for multi-domain proteins, surpassing most other prediction methods. Armadillo provides a simple, but effective method by which prediction of domain boundaries can be obtained with reasonable sensitivity. Armadillo should prove to be a valuable tool for rapidly delineating protein domains in poorly conserved proteins or those with no sequence neighbors. As a first-line predictor, domain meta-predictors could yield improved results with Armadillo predictions.     

Remote protein homology detection using recurrence quantification analysis and amino acid physicochemical properties

Remote homology detection refers to the detection of structure homology in evolutionarily related proteins with low sequence similarity. Supervised learning algorithms such as support vector machine (SVM) are currently the most accurate methods. In most of these SVM-based methods, efforts have been dedicated to developing new kernels to better use the pairwise alignment scores or sequence profiles. Moreover, amino acids’ physicochemical properties are not generally used in the feature representation of protein sequences. In this article, we present a remote homology detection method that incorporates two novel features: (1) a protein's primary sequence is represented using amino acid's physicochemical properties and (2) the similarity between two proteins is measured using recurrence quantification analysis (RQA). An optimization scheme was developed to select different amino acid indices (up to 10 for a protein family) that are best to characterize the given protein family. The selected amino acid indices may enable us to draw better biological explanation of the protein family classification problem than using other alignment-based methods. An SVM-based classifier will then work on the space described by the RQA metrics. The classification scheme is named as SVM-RQA. Experiments at the superfamily level of the SCOP1.53 dataset show that, without using alignment or sequence profile information, the features generated from amino acid indices are able to produce results that are comparable to those obtained by the published state-of-the-art SVM kernels. In the future, better prediction accuracies can be expected by combining the alignment-based features with our amino acids property-based features. Supplementary information including the raw dataset, the best-performing amino acid indices for each protein family and the computed RQA metrics for all protein sequences can be downloaded from http://ym151113.ym.edu.tw/svm-rqa.     

Identification and analysis of novel amino acid sequence repeats and domains in Pyrobaculum aerophilum using computational tools

We have identified four repeats and five domains that are novel in proteins encoded by the Pyrobaculum aerophilum str. IM2 proteome using automated in silico methods. A "repeat" corresponds to a region comprising less than 55 amino acid residues that occurs more than once in the protein sequence and sometimes present in tandem. A "domain" corresponds to a conserved region comprising greater than 55 amino acid residues and may be present as single or multiple copies in the protein sequence. These correspond to (1) 85 amino acid residues AAG domain, (2) 72 amino acid residues GFGN domain, (3) 43 amino acid residues KGG repeat, (4) 25 amino acid residues RWE repeat, (5) 25 amino acid residues RID repeat, (6) 108 amino acid residues NDFA domain, (7) 140 amino acid residues VxY domain, (8) 35 amino acid residues LLPN repeat and (9) 98 amino acid residues GxY domain. A repeat or domain is characterized by specific conserved sequence motifs. We discuss the presence of these repeats and domains in proteins from other genomes and their probable secondary structure.     

Toward a definition of the complete proteome of plant peroxisomes: Where experimental proteomics must be complemented by bioinformatics

In the past few years, proteome analysis of Arabidopsis peroxisomes has been established by the complementary efforts of four research groups and has emerged as the major unbiased approach to identify new peroxisomal proteins on a large scale. Collectively, more than 100 new candidate proteins from plant peroxisomes have been identified, including long-awaited low-abundance proteins. More than 50 proteins have been validated as peroxisome targeted, nearly doubling the number of established plant peroxisomal proteins. Sequence homologies of the new proteins predict unexpected enzyme activities, novel metabolic pathways and unknown non-metabolic peroxisome functions. Despite this remarkable success, proteome analyses of plant peroxisomes remain highly material intensive and require major preparative efforts. Characterization of the membrane proteome or post-translational protein modifications poses major technical challenges. New strategies, including quantitative mass spectrometry methods, need to be applied to allow further identifications of plant peroxisomal proteins, such as of stress-inducible proteins. In the long process of defining the complete proteome of plant peroxisomes, the prediction of peroxisome-targeted proteins from plant genome sequences emerges as an essential complementary approach to identify additional peroxisomal proteins that are, for instance, specific to peroxisome variants from minor tissues and organs or to abiotically stressed model and crop plants.     

Statistical analyses of protein sequence alignments identify structures and mechanisms in signal activation of sensor histidine kinases

Statistical analyses of genome sequence‐derived protein sequence data can identify amino acid residues that interact between proteins or between domains of a protein. These statistical methods are based on evolution‐directed amino acid variation responding to structural and functional constraints in proteins. The identified residues form a basis for determining structure and folding of proteins as well as inferring mechanisms of protein function. When applied to two‐component systems, several research groups have shown they can be used to identify the amino acid interactions between response regulators and histidine kinases and the specificity therein. Recently, statistical studies between the HisKA and HATPase‐ATP‐binding domains of histidine kinases identified amino acid interactions for both the inactive and the active catalytic states of such kinases. The identified interactions generated a model structure for the domain conformation of the active state. This conformation requires an unwinding of a portion of the C‐terminal helix of the HisKA domain that destroys the inactive state residue contacts and suggests how signal‐binding determines the equilibrium between the inactive and active states of histidine kinases. The rapidly accumulating protein sequence databases from genome, metagenome and microbiome studies are an important resource for functional and structural understanding of proteins and protein complexes in microbes.     

Prediction of protein domain boundaries from inverse covariances

It has been known even since relatively few structures had been solved that longer protein chains often contain multiple domains, which may fold separately and play the role of reusable functional modules found in many contexts. In many structural biology tasks, in particular structure prediction, it is of great use to be able to identify domains within the structure and analyze these regions separately. However, when using sequence data alone this task has proven exceptionally difficult, with relatively little improvement over the naive method of choosing boundaries based on size distributions of observed domains. The recent significant improvement in contact prediction provides a new source of information for domain prediction. We test several methods for using this information including a kernel smoothing‐based approach and methods based on building alpha‐carbon models and compare performance with a length‐based predictor, a homology search method and four published sequence‐based predictors: DOMCUT, DomPRO, DLP‐SVM, and SCOOBY‐DOmain. We show that the kernel‐smoothing method is significantly better than the other ab initio predictors when both single‐domain and multidomain targets are considered and is not significantly different to the homology‐based method. Considering only multidomain targets the kernel‐smoothing method outperforms all of the published methods except DLP‐SVM. The kernel smoothing method therefore represents a potentially useful improvement to ab initio domain prediction. Proteins 2013. © 2012 Wiley Periodicals, Inc.     

Identification of novel plant peroxisomal targeting signals by a combination of machine learning methods and in vivo subcellular targeting analyses

In the postgenomic era, accurate prediction tools are essential for identification of the proteomes of cell organelles. Prediction methods have been developed for peroxisome-targeted proteins in animals and fungi but are missing specifically for plants. For development of a predictor for plant proteins carrying peroxisome targeting signals type 1 (PTS1), we assembled more than 2500 homologous plant sequences, mainly from EST databases. We applied a discriminative machine learning approach to derive two different prediction methods, both of which showed high prediction accuracy and recognized specific targeting-enhancing patterns in the regions upstream of the PTS1 tripeptides. Upon application of these methods to the Arabidopsis thaliana genome, 392 gene models were predicted to be peroxisome targeted. These predictions were extensively tested in vivo, resulting in a high experimental verification rate of Arabidopsis proteins previously not known to be peroxisomal. The prediction methods were able to correctly infer novel PTS1 tripeptides, which even included novel residues. Twenty-three newly predicted PTS1 tripeptides were experimentally confirmed, and a high variability of the plant PTS1 motif was discovered. These prediction methods will be instrumental in identifying low-abundance and stress-inducible peroxisomal proteins and defining the entire peroxisomal proteome of Arabidopsis and agronomically important crop plants.     

Phylomat: an automated protein motif analysis tool for phylogenomics

Recent progress in genomics, proteomics, and bioinformatics enables unprecedented opportunities to examine the evolutionary history of molecular, cellular, and developmental pathways through phylogenomics. Accordingly, we have developed a motif analysis tool for phylogenomics (Phylomat, http://alg.ncsa.uiuc.edu/pmat) that scans predicted proteome sets for proteins containing highly conserved amino acid motifs or domains for in silico analysis of the evolutionary history of these motifs/domains. Phylomat enables the user to download results as full protein or extracted motif/domain sequences from each protein. Tables containing the percent distribution of a motif/domain in organisms normalized to proteome size are displayed. Phylomat can also align the set of full protein or extracted motif/domain sequences and predict a neighbor-joining tree from relative sequence similarity. Together, Phylomat serves as a user-friendly data-mining tool for the phylogenomic analysis of conserved sequence motifs/domains in annotated proteomes from the three domains of life.     

Protein Domain Analysis of Genomic Sequence Data Reveals Regulation of LRR Related Domains in Plant Transpiration in Ficus

Predicting protein domains is essential for understanding a protein’s function at the molecular level. However, up till now, there has been no direct and straightforward method for predicting protein domains in species without a reference genome sequence. In this study, we developed a functionality with a set of programs that can predict protein domains directly from genomic sequence data without a reference genome. Using whole genome sequence data, the programming functionality mainly comprised DNA assembly in combination with next-generation sequencing (NGS) assembly methods and traditional methods, peptide prediction and protein domain prediction. The proposed new functionality avoids problems associated with de novo assembly due to micro reads and small single repeats. Furthermore, we applied our functionality for the prediction of leucine rich repeat (LRR) domains in four species of Ficus with no reference genome, based on NGS genomic data. We found that the LRRNT_2 and LRR_8 domains are related to plant transpiration efficiency, as indicated by the stomata index, in the four species of Ficus. The programming functionality established in this study provides new insights for protein domain prediction, which is particularly timely in the current age of NGS data expansion.     

Enhanced structure prediction of gene products containing class III adenylyl cyclase domains

Domain finding algorithms are useful to understand overall domain architecture and to propose biological function to gene products. Automated methods of applying these tools to large-scale genome studies often employ stringent thresholds to recognize sequence domains. The realization of additional domains can be tedious involving manual intervention but can lead to better understanding of overall biological function. We propose a multi-step approach for the further examination of unassigned linker regions that exploits properties such as the conservation of domain architectures of homologous proteins to propose connections. Improved structure prediction is possible starting from initial domain architectures, obtained from simple 'domain finding' techniques, by concentrating on connecting unassigned regions. 254 unassigned regions have been examined in 114 gene products that potentially contain at least one class III adenylyl cyclase domain for a pilot study. Reliable structure prediction was possible for nearly 80% of unassigned regions. New connections were recognized that assign putative structure and function to these regions by indirect searches (26%). Several others (34%) could be associated with three-dimensional models that might pertain to novel folds and new functions with enough structural content and evolutionary conservation. The presence of additional domains will provide further clues to the overall function of the gene products and their recruitment in particular biochemical pathways.     

The role of internal duplication in the evolution of multi-domain proteins

Many proteins consist of several structural domains. These multi-domain proteins have likely been generated by selective genome growth dynamics during evolution to perform new functions as well as to create structures that fold on a biologically feasible time scale. Domain units frequently evolved through a variety of genetic shuffling mechanisms. Here we examine the protein domain statistics of more than 1000 organisms including eukaryotic, archaeal and bacterial species. The analysis extends earlier findings on asymmetric statistical laws for proteome to a wider variety of species. While proteins are composed of a wide range of domains, displaying a power-law decay, the computation of domain families for each protein reveals an exponential distribution, characterizing a protein universe composed of a thin number of unique families. Structural studies in proteomics have shown that domain repeats, or internal duplicated domains, represent a small but significant fraction of genome. In spite of its importance, this observation has been largely overlooked until recently. We model the evolutionary dynamics of proteome and demonstrate that these distinct distributions are in fact rooted in an internal duplication mechanism. This process generates the contemporary protein structural domain universe, determines its reduced thickness, and tames its growth. These findings have important implications, ranging from protein interaction network modeling to evolutionary studies based on fundamental mechanisms governing genome expansion.