DeepNphos: A deep-learning architecture for prediction of N-phosphorylation sites

MOTIVATION

Phosphorylation, a prevalent post-translational modification, plays a crucial role in regulating cellular activities. This process encompasses O-phosphorylation (e.g., phosphoserine) and N-phosphorylation (e.g., phospho-lysine (pK), phospho-arginine (pR), and phospho-histidine (pH)). While significant research has focused on O-phosphorylation, resulting in the development of various algorithms for predicting O-phosphorylation sites with commendable performance, there has been a notable absence of models designed to predict N-phosphorylation sites. This study introduces an integrated model named DeepNphos, designed to predict N-phosphorylation sites. This model is developed based on the analysis of thousands of experimentally identified pK, pR and pH sites.

RESULTS

Observing that the Convolutional Neural Network (CNN) model, incorporating the One-Hot encoding feature, demonstrates favorable performance in comparison to other models when predicting pK, pR, and pH sites. Additionally, pK exhibits similarities to other lysine modification types, and integrating the CNN model with a deep-transfer learning (DTL) strategy based on tens of thousands of known lysine modification sites could enhance pK prediction performance. In contrast, pR exhibits little similarity to other arginine modification types, and the integration of DTL has minimal impact on pR prediction performance. Furthermore, the decision was made to refrain from incorporating the DTL strategy in predicting pH sites, given the scarcity of histidine modification sites beyond those associated with pH. The final classifiers for predicting pK, pR, and pH sites achieve AUC values of 0.856, 0.805 and 0.802 for ten-fold cross-validation, respectively. Overall, DeepNphos is the first classifier for predicting N-phosphorylation sites, accessible at https://github.com/ChangXulinmessi/DeepNPhos.

pHisPred: a tool for the identification of histidine phosphorylation sites by integrating amino acid patterns and properties

Background

Protein histidine phosphorylation (pHis) plays critical roles in prokaryotic signal transduction pathways and various eukaryotic cellular processes. It is estimated to account for 6–10% of the phosphoproteome, however only hundreds of pHis sites have been discovered to date. Due to the inherent disadvantages of experimental methods, it is an urgent task for developing efficient computational approaches to identify pHis sites.

Results

Here, we present a novel tool, pHisPred, for accurately identifying pHis sites from protein sequences. We manually collected the largest number of experimental validated pHis sites to build benchmark datasets. Using randomized tenfold CV, the weighted SVM-RBF model shows the best performance than other four commonly used classification models (LR, KNN, RF, and MLP). From ten thousands of features, 140 and 150 most informative features were individually selected out for eukaryotic and prokaryotic models. The average AUC and F1-score values of pHisPred were (0.81, 0.40) and (0.78, 0.46) for tenfold CV on the eukaryotic and prokaryotic training datasets, respectively. In addition, pHisPred significantly outperforms other tools on testing datasets, in particular on the eukaryotic one.

Conclusion

We implemented a python program of pHisPred, which is freely available for non-commercial use at https://github.com/xiaofengsong/pHisPred. Moreover, users can use it to train new models with their own data.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04938-x.

HisPhosSite: A comprehensive database of histidine phosphorylated proteins and sites

Histidine phosphorylation is critically important in a variety of cellular processes including signal transduction, cell cycle, proliferation, differentiation, and apoptosis. It is estimated to account for 6% of all phosphorylated amino acids. However, due to the acid lability of the PN bond, the study of pHis lags far behind that of pSer, pThr, and pTyr. Recently, the development and use of pHis-specific antibodies and methodologies have led to a resurgence in the study of histidine phosphorylation. Although a considerable number of pHis proteins and sites have been discovered, most of them have not been manually curated and integrated to any databases. There is a lack of a data repository for pHis, and such work is expected to help further systemic studies of pHis. Thus, we present a comprehensive resource database of histidine phosphorylation (HisPhosSite) by curating experimentally validated pHis proteins and sites and compiling putative pHis sites with ortholog search. HisPhosSite contains 776 verified pHis sites and 2702 verified pHis proteins in 38 eukaryotic and prokaryotic species and 15,378 putative pHis sites and 10,816 putative pHis proteins in 1366 species. HisPhosSite provides rich annotations of pHis sites and proteins and multiple search engines (including motif search and BLAST search) for users to locate pHis sites of interest. HisPhosSite is available at http://reprod.njmu.edu.cn/hisphossite. SIGNIFICANCE: Histidine phosphorylation is involved in a variety of cellular processes as well as cancers, and it has been proved to be more common than previously thought. The HisPhosSite database was developed to collect pHis data from published literatures with experimental evidences. Unification of the identified pHis proteins and sites will give researchers an informative resource for histidine phosphorylation. HisPhosSite has a user-friendly interface with multiple search engines for users to locate pHis sites of interest. In addition, the database provides rich structural and functional annotations. HisPhosSite will help future studies and elucidation of the functions of histidine phosphorylation.

Histidine kinase NME1 and NME2 are involved in TGF-β1-induced HSC activation and CCl4-induced liver fibrosis

Histidine phosphorylation (pHis) was first reported in 1962. There are few studies on pHis because of the thermal and acidic instability of pHis and the lack of specific methods to detect it. pHis has two isomers of 1-phosphate histidine (1-pHis) and 3-phosphate histidine (3-pHis). pHis antibodies have been developed recently and have promoted research in this field. In this study, we established a CCl₄-induced liver fibrosis model in C57 mice and a TGF-β1-induced HSC activation model in LX-2 cells, to study the role of histidine phosphorylation. The expression of histidine kinases NME1 and NME2 was increased, histidine phosphatase PGAM5 and PHPT1 was unchanged, and 1-pHis and 3-pHis were increased in the in vivo and in vitro models. The expression of LHPP was decreased in the in vivo model but not in the in vitro model. To further study the role of NME1, NME2, and histidine phosphorylation in HSC activation, we silenced NME1 or NME2 and administered TGF-β1 in LX-2 cells. The results showed silencing NME1 or NME2 decreased TGF-β1-induced pHis levels and the expression of α-SMA and COL1A1, indicating the activation of HSC was suppressed. Then, we found the inhibitory effect on HSC activation is due to reduced phosphorylation of Smad2 and Smad3. In summary, our studies indicate that NME1 and NME2 are involved in TGF-β1-induced HSC activation and CCl₄-induced liver fibrosis, which may be mediated by histidine phosphorylation.

Subcellular Localization of Histidine Phosphorylated Proteins Through Indirect Immunofluorescence

Immunofluorescence (IF) takes advantage of biological and physical mechanisms to identify proteins in cell or tissue samples, exploiting the specificity of antibodies and stimulated fluorescence light emission. Here, we describe an immunofluorescence staining method for the identification of histidine phosphorylated proteins that uses neutral/alkaline conditions and targeted reagents to overcome the chemical lability of histidine phosphorylation. This method describes how 1- and 3-phosphohistidine (pHis) monoclonal antibodies can be used to reveal the localization of proteins containing these elusive phosphoramidate bonds in cells. Standard procedures and materials for IF staining with adherent and nonadherent cells are described.

Determination of Phosphohistidine Stoichiometry in Histidine Kinases by Intact Mass Spectrometry

Protein histidine phosphorylation has largely remained unexplored due to the challenges of analyzing relatively unstable phosphohistidine-containing proteins. We describe a procedure for determining the stoichiometry of histidine phosphorylation on the human histidine kinases NME1 and NME2 by intact mass spectrometry under conditions that retain this acid-labile protein modification. By characterizing these two model histidine protein kinases in the absence and presence of a suitable phosphate donor, the stoichiometry of histidine phosphorylation can be determined. The described method can be readily adapted for the analysis of other proteins containing phosphohistidine.

Empirical Evidence of Cellular Histidine Phosphorylation by Immunoblotting Using pHis mAbs

Immunoblotting is a ubiquitous immunological technique that aids in detecting and quantifying proteins (including those of lower abundance) and their posttranslational modifications such as phosphorylation, acetylation, ubiquitylation, and sumoylation. The technique involves electrophoretically separating proteins on an SDS-PAGE gel, transferring them onto a PVDF (or nitrocellulose) membrane and probing with specific antibodies. Here we describe an immunoblotting technique for detecting cellular phosphohistidine, a labile posttranslational modification, by optimizing experimental conditions such that the labile phosphohistidine signal is conserved throughout the experiment.

Immunohistochemistry (IHC): Chromogenic Detection of 3-Phosphohistidine Proteins in Formaldehyde-Fixed, Frozen Mouse Liver Tissue Sections

The development of antibodies that specifically detect histidine-phosphorylated proteins is a recent achievement and allows potential roles of histidine phosphorylated proteins in pathological and physiological conditions to be characterized. Immunohistochemical analyses enable the detection of proteins in tissues and can reveal alterations to the quantity and/or localization of these proteins through comparisons of normal and diseased specimens. However, the sensitivity of phosphohistidine modifications to phosphatases, acidic pH, and elevated temperatures poses unique challenges to the detection process and requires a protocol that bypasses traditional procedures utilizing paraffin-embedding and antigen-retrieval methods. Here, we detail a method for a brief fixation by 4% (v/v) paraformaldehyde on freshly collected tissues in the presence of PhosSTOP to block phosphatase activity, followed by a float on sucrose to protect the tissue prior to freezing. Specimens are then embedded in a cryopreservation medium in molds and frozen using an isoflurane, dry ice bath to best preserve the tissue morphology and phosphohistidine signal. We validate this technique in normal mouse liver using SC44-1, a monoclonal anti-3-pHis antibody used to uncover a role for a protein histidine phosphatase as a tumor suppressor in the liver. Furthermore, we demonstrate that the antibody signal can be eliminated by preincubating SC44-1 with a peptide treated with phosphoramidate to phosphorylate histidine residues. Thus, we present an IHC protocol suitable for specific detection of 3-phosphohistidine proteins in mouse liver tissue, and suggest that this can be used as a starting point for optimization of IHC using other phosphohistidine antibodies or in other tissue types, generating information that will enhance our understanding of phosphohistidine in models of disease.

Analysis of 1- and 3-Phosphohistidine (pHis) Protein Modification Using Model Enzymes Expressed in Bacteria

Despite the discovery of protein histidine (His) phosphorylation nearly six decades ago, difficulties in measuring and quantifying this unstable post-translational modification (PTM) have limited its mechanistic analysis in prokaryotic and eukaryotic signaling. Here, we describe reliable procedures for affinity purification, cofactor-binding analysis and antibody-based detection of phosphohistidine (pHis), on the putative human His kinases NME1 (NDPK-A) and NME2 (NDPK-B) and the glycolytic phosphoglycerate mutase PGAM1. By exploiting isomer-specific monoclonal N1-pHis and N3-pHis antibodies, we describe robust protocols for immunological detection and isomer discrimination of site-specific pHis, including N3-pHis on His 11 of PGAM1.

Electron Transfer/Higher Energy Collisional Dissociation of Doubly Charged Peptide Ions: Identification of Labile Protein Phosphorylations

In recent years, labile phosphorylation sites on arginine, histidine, cysteine, and lysine as well as pyrophosphorylation of serine and threonine have gained more attention in phosphoproteomic studies. However, the analysis of these delicate posttranslational modifications via tandem mass spectrometry remains a challenge. Common fragmentation techniques such as collision-induced dissociation (CID) and higher energy collisional dissociation (HCD) are limited due to extensive phosphate-related neutral loss. Electron transfer dissociation (ETD) has shown to preserve labile modifications, but is restricted to higher charge states, missing the most prevalent doubly charged peptides. Here, we report the ability of electron transfer/higher energy collisional dissociation (EThcD) to fragment doubly charged phosphorylated peptides without losing the labile modifications. Using synthetic peptides that contain phosphorylated arginine, histidine, cysteine, and lysine as well as pyrophosphorylated serine residues, we evaluated the optimal fragmentation conditions, demonstrating that EThcD is the method of choice for unambiguous assignment of tryptic, labile phosphorylated peptides. Graphical Abstract.

Site-specific His/Asp phosphoproteomic analysis of prokaryotes reveals putative targets for drug resistance

BACKGROUND

Phosphorylation of amino acid residues on proteins is an important and common post-translational modification in both eukaryotes and prokaryotes. Most research work has been focused on phosphorylation of serine, threonine or tyrosine residues, whereas phosphorylation of other amino acids are significantly less clear due to the controversy on their stability under standard bioanalytical conditions.

RESULTS

Here we applied a shotgun strategy to analyze the histidine and aspartate phosphorylations in different microbes. Our results collectively indicate that histidine and aspartate phosphorylations frequently occur also in proteins that are not part of the two-component systems. Noticeably, a number of the modified proteins are pathogenesis-related or essential for survival in host. These include the zinc ion periplasmic transporter ZnuA in Acinetobacter baumannii SK17, the multidrug and toxic compound extrusion (MATE) channel YeeO in Klebsiella pneumoniae NTUH-K2044, branched amino acid transporter AzlC in Vibrio vulnificus and the RNA-modifying pseudouridine synthase in Helicobacter pylori.

CONCLUSIONS

In summary, histidine and aspartate phosphorylation is likely to be ubiquitous and to take place in proteins of various functions. This work also sheds light into how these functionally important proteins and potential drug targets might be regulated at a post-translational level.

Gas-phase intermolecular phosphate transfer within a phosphohistidine phosphopeptide dimer

The hydrogen bonds and electrostatic interactions that form between the protonated side chain of a basic residue and the negatively charged phosphate of a phosphopeptide can play crucial roles in governing their dissociation pathways under low-energy collision-induced dissociation (CID). Understanding how phosphoramidate (i.e. phosphohistidine, phospholysine and phosphoarginine), rather than phosphomonoester-containing peptides behave during CID is paramount in investigation of these problematic species by tandem mass spectrometry. To this end, a synthetic peptide containing either phosphohistidine (pHis) or phospholysine (pLys) was analyzed by ESI-MS using a Paul-type ion trap (AmaZon, Bruker) and by traveling wave ion mobility-mass spectrometry (Synapt G2-Si, Waters). Analysis of the products of low-energy CID demonstrated formation of a doubly 'phosphorylated' product ion arising from intermolecular gas-phase phosphate transfer within a phosphopeptide dimer. The results are explained by the formation of a homodimeric phosphohistidine (pHis) peptide non-covalent complex (NCX), likely stabilized by the electrostatic interaction between the pHis phosphate group and the protonated C-terminal lysine residue of the peptide. To the best of our knowledge this is the first report of intermolecular gas-phase phosphate transfer from one phosphopeptide to another, leading to a doubly phosphorylated peptide product ion.