Identification of the sequence determinants of protein N-terminal acetylation through a decision tree approach

Prediction of protein N-terminal acetylation modification sites based on CNN-BiLSTM-attention model

N-terminal acetylation is one of the most common and important post-translational modifications (PTM) of eukaryotic proteins. PTM plays a crucial role in various cellular processes and disease pathogenesis. Thus, the accurate identification of N-terminal acetylation modifications is important to gain insight into cellular processes and other possible functional mechanisms. Although some algorithmic models have been proposed, most have been developed based on traditional machine learning algorithms and small training datasets. Their practical applications are limited. Nevertheless, deep learning algorithmic models are better at handling high-throughput and complex data. In this study, DeepCBA, a model based on the hybrid framework of convolutional neural network (CNN), bidirectional long short-term memory network (BiLSTM), and attention mechanism deep learning, was constructed to detect the N-terminal acetylation sites. The DeepCBA was built as follows: First, a benchmark dataset was generated by selecting low-redundant protein sequences from the Uniport database and further reducing the redundancy of the protein sequences using the CD-HIT tool. Subsequently, based on the skip-gram model in the word2vec algorithm, tripeptide word vector features were generated on the benchmark dataset. Finally, the CNN, BiLSTM, and attention mechanism were combined, and the tripeptide word vector features were fed into the stacked model for multiple rounds of training. The model performed excellently on independent dataset test, with accuracy and area under the curve of 80.51% and 87.36%, respectively. Altogether, DeepCBA achieved superior performance compared with the baseline model, and significantly outperformed most existing predictors. Additionally, our model can be used to identify disease loci and drug targets.

The role of N-acetyltransferases in cancers

Cancer is a major global health problem that disrupts the balance of normal cellular growth and behavior. Mounting evidence has shown that epigenetic modification, specifically N-terminal acetylation, play a crucial role in the regulation of cell growth and function. Acetylation is a co- or post-translational modification to regulate important cellular progresses such as cell proliferation, cell cycle progress, and energy metabolism. Recently, N-acetyltransferases (NATs), enzymes responsible for acetylation, regulate signal transduction pathway in various cancers including hepatocellular carcinoma, breast cancer, lung cancer, colorectal cancer and prostate cancer. In this review, we clarify the regulatory role of NATs in cancer progression, such as cell proliferation, metastasis, cell apoptosis, autophagy, cell cycle arrest and energy metabolism. Furthermore, the mechanism of NATs on cancer remains to be further studied, and few drugs have been developed. This provides us with a new idea that targeting acetylation, especially NAT-mediated acetylation, may be an attractive way for inhibiting cancer progression.

VirulentPred 2.0: An improved method for prediction of virulent proteins in bacterial pathogens

Virulence proteins in pathogens are essential for causing disease in a host. They enable the pathogen to invade, survive and multiply within the host, thus enhancing its potential to cause disease while also causing evasion of host defense mechanisms. Identifying these factors, especially potential vaccine candidates or drug targets, is critical for vaccine or drug development research. In this context, we present an improved version of VirulentPred 1.0 for rapidly identifying virulent proteins. The VirulentPred 2.0 is based on training machine learning models with experimentally validated virulent protein sequences. VirulentPred 2.0 achieved 84.71% accuracy with the validation dataset and 85.18% on an independent test dataset. The models are trained and evaluated with the latest sequence datasets of virulent proteins, which are three times greater in number than the proteins used in the earlier version of VirulentPred. Moreover, a significant improvement of 11% in the prediction accuracy over the earlier version is achieved with the best position-specific scoring matrix (PSSM)-based model for the latest test dataset. VirulentPred 2.0 is available as a user-friendly web interface at https://bioinfo.icgeb.res.in/virulent2/ and a standalone application suitable for bulk predictions. With higher efficiency and availability as a standalone tool, VirulentPred 2.0 holds immense potential for high throughput yet efficient identification of virulent proteins in bacterial pathogens.

N-terminal acetyltransferase NatB regulates Rad51-dependent repair of double-strand breaks in Saccharomyces cerevisiae

Homologous recombination (HR) is a highly accurate mechanism for repairing DNA double-strand breaks (DSBs) that arise from various genotoxic insults and blocked replication forks. Defects in HR and unscheduled HR can interfere with other cellular processes such as DNA replication and chromosome segregation, leading to genome instability and cell death. Therefore, the HR process has to be tightly controlled. Protein N-terminal acetylation is one of the most common modifications in eukaryotic organisms. Studies in budding yeast implicate a role for NatB acetyltransferase in HR repair, but precisely how this modification regulates HR repair and genome integrity is unknown. In this study, we show that cells lacking NatB, a dimeric complex composed of Nat3 and Mdm2, are sensitive to the DNA alkylating agent methyl methanesulfonate (MMS), and that overexpression of Rad51 suppresses the MMS sensitivity of nat3Δ cells. Nat3-deficient cells have increased levels of Rad52-yellow fluorescent protein foci and fail to repair DSBs after release from MMS exposure. We also found that Nat3 is required for HR-dependent gene conversion and gene targeting. Importantly, we observed that nat3Δ mutation partially suppressed MMS sensitivity in srs2Δ cells and the synthetic sickness of srs2Δ sgs1Δ cells. Altogether, our results indicate that NatB functions upstream of Srs2 to activate the Rad51-dependent HR pathway for DSB repair.

NAA60 (HAT4): the newly discovered bi-functional Golgi member of the acetyltransferase family

Chromatin structural organization, gene expression and proteostasis are intricately regulated in a wide range of biological processes, both physiological and pathological. Protein acetylation, a major post-translational modification, is tightly involved in interconnected biological networks, modulating the activation of gene transcription and protein action in cells. A very large number of studies describe the pivotal role of the so-called acetylome (accounting for more than 80% of the human proteome) in orchestrating different pathways in response to stimuli and triggering severe diseases, including cancer. NAA60/NatF (N-terminal acetyltransferase F), also named HAT4 (histone acetyltransferase type B protein 4), is a newly discovered acetyltransferase in humans modifying N-termini of transmembrane proteins starting with M-K/M-A/M-V/M-M residues and is also thought to modify lysine residues of histone H4. Because of its enzymatic features and unusual cell localization on the Golgi membrane, NAA60 is an intriguing acetyltransferase that warrants biochemical and clinical investigation. Although it is still poorly studied, this review summarizes current findings concerning the structural hallmarks and biological role of this novel targetable epigenetic enzyme.

N-terminomics - its past and recent advancements

N-terminomics is a rapidly evolving branch of proteomics that encompasses the study of protein N-terminal sequence. A proteome-wide collection of such sequences has been widely used to understand the proteolytic cascades and in annotating the genome. Over the last two decades, various N-terminomic strategies have been developed for achieving high sensitivity, greater depth of coverage, and high-throughputness. We, in this review, cover how the field of N-terminomics has evolved to date, including discussion on various sample preparation and N-terminal peptide enrichment strategies. We also compare different N-terminomic methods and highlight their relative benefits and shortcomings in their implementation. In addition, an overview of the currently available bioinformatics tools and data analysis pipelines for the annotation of N-terminomic datasets is also included. SIGNIFICANCE: It has been recognized that proteins undergo several post-translational modifications (PTM), and a number of perturbed biological pathways are directly associated with modifications at the terminal sites of a protein. In this regard, N-terminomics can be applied to generate a proteome-wide landscape of mature N-terminal sequences, annotate their source of generation, and recognize their significance in the biological pathways. Besides, a system-wide study can be used to study complicated proteolytic machinery and protease cleavage patterns for potential therapeutic targets. Moreover, due to unprecedented improvements in the analytical methods and mass spectrometry instrumentation in recent times, the N-terminomic methodologies now offers an unparalleled ability to study proteoforms and their implications in clinical conditions. Such approaches can further be applied for the detection of low abundant proteoforms, annotation of non-canonical protein coding sites, identification of candidate disease biomarkers, and, last but not least, the discovery of novel drug targets.

A note on the interpretation of tree-based regression models

Tree-based models are a popular tool for predicting a response given a set of explanatory variables when the regression function is characterized by a certain degree of complexity. Sometimes, they are also used to identify important variables and for variable selection. We show that if the generating model contains chains of direct and indirect effects, then the typical variable importance measures suggest selecting as important mainly the background variables, which have a strong indirect effect, disregarding the variables that directly influence the response. This is attributable mainly to the variable choice in the first steps of the algorithm selecting the splitting variable and to the greedy nature of such search. This pitfall could be relevant when using tree-based algorithms for understanding the underlying generating process, for population segmentation and for causal inference.

In Vitro N-Terminal Acetylation of Bacterially Expressed Parvalbumins by N-Terminal Acetyltransferases from Escherichia coli

Most eukaryotic proteins are N-terminally acetylated (Nt-acetylated) by specific N-terminal acetyltransferases (NATs). Although this co-/post-translational protein modification may affect different aspects of protein functioning, it is typically neglected in studies of bacterially expressed eukaryotic proteins, lacking this modification. To overcome this limitation of bacterial expression, we have probed the efficiency of recombinant Escherichia coli NATs (RimI, RimJ, and RimL) with regard to in vitro Nt-acetylation of several parvalbumins (PAs) expressed in E. coli. PA is a calcium-binding protein of vertebrates, which is sensitive to Nt-acetylation. Our analyses revealed that only metal-free PAs were prone to Nt-acetylation (up to 100%), whereas Ca²⁺ binding abolished this modification, thereby indicating that Ca²⁺-induced structural stabilization of PAs impedes their Nt-acetylation. RimJ and RimL were active towards all PAs with N-terminal serine. Their activity towards PAs beginning with alanine was PA-specific, suggesting the importance of the subsequent residues. RimI showed the least activity regardless of the PA studied. Overall, NATs from E. coli are suited for post-translational Nt-acetylation of bacterially expressed eukaryotic proteins with decreased structural stability.

Prediction of Extracellular Matrix Proteins by Fusing Multiple Feature Information, Elastic Net, and Random Forest Algorithm

Extracellular matrix (ECM) proteins play an important role in a series of biological processes of cells. The study of ECM proteins is helpful to further comprehend their biological functions. We propose ECMP-RF (extracellular matrix proteins prediction by random forest) to predict ECM proteins. Firstly, the features of the protein sequence are extracted by combining encoding based on grouped weight, pseudo amino-acid composition, pseudo position-specific scoring matrix, a local descriptor, and an autocorrelation descriptor. Secondly, the synthetic minority oversampling technique (SMOTE) algorithm is employed to process the class imbalance data, and the elastic net (EN) is used to reduce the dimension of the feature vectors. Finally, the random forest (RF) classifier is used to predict the ECM proteins. Leave-one-out cross-validation shows that the balanced accuracy of the training and testing datasets is 97.3% and 97.9%, respectively. Compared with other state-of-the-art methods, ECMP-RF is significantly better than other predictors.

A novel molecular representation with BiGRU neural networks for learning atom

Molecular representations play critical roles in researching drug design and properties, and effective methods are beneficial to assisting in the calculation of molecules and solving related problem in drug discovery. In previous years, most of the traditional molecular representations are based on hand-crafted features and rely heavily on biological experimentations, which are often costly and time consuming. However, recent researches achieve promising results using machine learning on various domains. In this article, we present a novel method named Smi2Vec-BiGRU that is designed for learning atoms and solving the single- and multitask binary classification problems in the field of drug discovery, which are the basic and also key problems in this field. Specifically, our approach transforms the molecule data in the SMILES format into a set of sample vectors and then feeds them into the bidirectional gated recurrent unit neural networks for training, which learns low-dimensional vector representations for molecular drug. We conduct extensive experiments on several widely used benchmarks including Tox21, SIDER and ClinTox. The experimental results show that our approach can achieve state-of-the-art performance on these benchmarking datasets, demonstrating the feasibility and competitiveness of our proposed approach.

N-acetyltransferases from three different organisms displaying distinct selectivity toward hexosamines and N-terminal amine of peptides

N-acetyltransferases are a family of enzymes that catalyze the transfer of the acetyl moiety (COCH₃) from acetyl coenzyme A (Acetyl-CoA) to a primary amine of acceptor substrates from small molecules such as aminoglycoside to macromolecules of various proteins. In this study, the substrate selectivity of three N-acetyltransferases falling into different phylogenetic groups was probed against a series of hexosamines and synthetic peptides. GlmA from Clostridium acetobutylicum and RmNag from Rhizomucor miehei, which have been defined as glucosamine N-acetyltransferases, were herein demonstrated to be also capable of acetylating the free amino group on the very first glycine residue of peptide in spite of varied catalytic efficiency. The human recombinant N-acetyltransferase of Naa10p, however, prefers primary amine groups in the peptides as opposed to glucosamine. The varied preference of GlmA, RmNag and Naa10p probably arose from the divergent evolution of these N-acetyltransferases. The expanded knowledge of acceptor specificity would as well facilitate the application of these N-acetyltransferases in the acetylation of hexosamines or peptides.