FuncPhos-STR: An integrated deep neural network for functional phosphosite prediction based on AlphaFold protein structure and dynamics.

Phosphorylation modifications play important regulatory roles in most biological processes. However, the functional assignment for the vast majority of the identified phosphosites remains a major challenge. Here, we provide a deep learning framework named FuncPhos-STR as an online resource, for functional prediction and structural visualization of human proteome-level phosphosites. Based on our reported FuncPhos-SEQ framework, which was built by integrating phosphosite sequence evolution and protein-protein interaction (PPI) information, FuncPhos-STR was developed by further integrating the structural and dynamics information on AlphaFold protein structures. The characterized structural topology and dynamics features underlying functional phosphosites emphasized their molecular mechanism for regulating protein functions. By integrating the structural and dynamics, sequence evolutionary, and PPI network features from protein different dimensions, FuncPhos-STR has advantage over other reported models, with the best AUC value of 0.855. Using FuncPhos-STR, the phosphosites inside the pocket regions are accessible to higher functional scores, theoretically supporting their potential regulatory mechanism. Overall, FuncPhos-STR would accelerate the functional identification of huge unexplored phosphosites, and facilitate the elucidation of their allosteric regulation mechanisms. The web server of FuncPhos-STR is freely available at http://funcptm.jysw.suda.edu.cn/str.

Deciphering the functional landscape of phosphosites with deep neural network.

Current biochemical approaches have only identified the most well-characterized kinases for a tiny fraction of the phosphoproteome, and the functional assignments of phosphosites are almost negligible. Herein, we analyze the substrate preference catalyzed by a specific kinase and present a novel integrated deep neural network model named FuncPhos-SEQ for functional assignment of human proteome-level phosphosites. FuncPhos-SEQ incorporates phosphosite motif information from a protein sequence using multiple convolutional neural network (CNN) channels and network features from protein-protein interactions (PPIs) using network embedding and deep neural network (DNN) channels. These concatenated features are jointly fed into a heterogeneous feature network to prioritize functional phosphosites. Combined with a series of in vitro and cellular biochemical assays, we confirm that NADK-S48/50 phosphorylation could activate its enzymatic activity. In addition, ERK1/2 are discovered as the primary kinases responsible for NADK-S48/50 phosphorylation. Moreover, FuncPhos-SEQ is developed as an online server.

Elastic network models and molecular dynamic simulations reveal the molecular basis of allosteric regulation in ubiquitin-specific protease 7 (USP7).

Ubiquitin-specific protease 7 (USP7) is one of the most abundant deubiquitinases and plays an important role in various malignant tumors. However, the molecular mechanisms underlying USP7's structures, dynamics, and biological significance are yet to be investigated. In this study, we constructed the full-length models of USP7 in both the extended and compact state, and applied elastic network models (ENM), molecular dynamics (MD) simulations, perturbation response scanning (PRS) analysis, residue interaction networks as well as allosteric pocket prediction to investigate allosteric dynamics in USP7. Our analysis of intrinsic and conformational dynamics revealed that the structural transition between the two states is characterized by global clamp motions, during which the catalytic domain (CD) and UBL4-5 domain exhibit strong negative correlations. The PRS analysis, combined with the analysis of disease mutations and post-translational modifications (PTMs) further highlighted the allosteric potential of the two domains. The residue interaction network based on MD simulations captured an allosteric communication path which starts at CD domain and ends at UBL4-5 domain. Moreover, we identified a pocket at the TRAF-CD interface as a high-potential allosteric site for USP7. Overall, our studies not only provide molecular insights into the conformational changes of USP7, but also aid in the design of allosteric modulators that target USP7.

Protectin DX Relieve Hyperoxia-induced Lung Injury by Protecting Pulmonary Endothelial Glycocalyx.

Background: Bronchopulmonary dysplasia (BPD) is a common chronic lung disease in premature infants with limited treatments and poor prognosis. Damaged endothelial glycocalyx leads to vascular permeability, lung edema and inflammation. However, whether hyperoxia increases neonatal pulmonary microvascular permeability by degrading the endothelial glycocalyx remains unknown. Methods: Newborn mice were maintained in 60-70% O2 for 7 days. Protectin DX (PDX), an endogenous lipid mediator, was injected intraperitoneally on postnatal d 0, 2, 4 and 6. Lung samples and bronchoalveolar lavage fluid were taken at the end of the study. Primary human umbilical vein endothelial cells (HUVECs) were cultured in 80%O2. Results: Hyperoxia exposure for 7 days led to neonatal mice alveolar simplification with less radial alveolar count (RAC), mean linear intercept (MLI) and mean alveolar diameter (MAD) compared to the control group. Hyperoxia exposure increased lung vascular permeability with more fluid and proteins and inflammatory factors, including TNF-α and IL-1ß, in bronchoalveolar lavage fluid while reducing the heparan sulfate (HS), the most abundant component of the endothelial glycocalyx, in the pulmonary endothelial cells. PDX relieve these changes. PDX attenuated hyperoxia-induced high expression of heparanase (HPA), the endoglycosidase that shed endothelial glycocalyx, p-P65, P65, and low expression of SIRT1. BOC-2 and EX527 abolished the affection of PDX both in vivo and intro. Conclusion: In summary, our findings indicate that PDX treatment relieves hyperoxia-induced alveolar simplification, vascular leakage and lung inflammation by attenuating pulmonary endothelial glycocalyx injury via the SIRT1/NF-κB/ HPA pathway.

A RhoA structure with switch II flipped outward revealed the conformational dynamics of switch II region.

Small GTPase RhoA switches from GTP-bound state to GDP-bound state by hydrolyzing GTP, which is accelerated by GTPases activating proteins (GAPs). However, less study of RhoA structural dynamic changes was conducted during this process, which is essential for understanding the molecular mechanism of GAP dissociation. Here, we solved a RhoA structure in GDP-bound state with switch II flipped outward. Because lacking the intermolecular interactions with guanine nucleotide, we proposed this conformation of RhoA could be an intermediate after GAP dissociation. Further molecular dynamics simulations found the conformational changes of switch regions are indeed existing in RhoA and involved in the regulation of GAP dissociation and GEF recognition. Besides, the guanine nucleotide binding pocket extended to switch II region, indicating a potential "druggable" cavity for RhoA. Taken together, our study provides a deeper understanding of the dynamic properties of RhoA switch regions and highlights the direction for future drug development.

PPICT: an integrated deep neural network for predicting inter-protein PTM cross-talk.

Post-translational modifications (PTMs) fine-tune various signaling pathways not only by the modification of a single residue, but also by the interplay of different modifications on residue pairs within or between proteins, defined as PTM cross-talk. As a challenging question, less attention has been given to PTM dynamics underlying cross-talk residue pairs and structural information underlying protein-protein interaction (PPI) graph, limiting the progress in this PTM functional research. Here we propose a novel integrated deep neural network PPICT (Predictor for PTM Inter-protein Cross-Talk), which predicts PTM cross-talk by combining protein sequence-structure-dynamics information and structural information for PPI graph. We find that cross-talk events preferentially occur among residues with high co-evolution and high potential in allosteric regulation. To make full use of the complex associations between protein evolutionary and biophysical features, and protein pair features, a heterogeneous feature combination net is introduced in the final prediction of PPICT. The comprehensive test results show that the proposed PPICT method significantly improves the prediction performance with an AUC value of 0.869, outperforming the existing state-of-the-art methods. Additionally, the PPICT method can capture the potential PTM cross-talks involved in the functional regulatory PTMs on modifying enzymes and their catalyzed PTM substrates. Therefore, PPICT represents an effective tool for identifying PTM cross-talk between proteins at the proteome level and highlights the hints for cross-talk between different signal pathways introduced by PTMs.

HNRNPA2B1 as a potential therapeutic target for thymic epithelial tumor recurrence: An integrative network analysis.

Thymic epithelial tumors (TETs) are rare malignant tumors, and the molecular mechanisms of both primary and recurrent TETs are poorly understood. Here we established comprehensive proteomic signatures of 15 tumors (5 recurrent and 10 non-recurrent) and 15 pair wised tumor adjacent normal tissues. We then proposed an integrative network approach for studying the proteomics data by constructing protein-protein interaction networks based on differentially expressed proteins and a machine learning-based score, followed by network modular analysis, functional enrichment annotation and shortest path inference analysis. Network modular analysis revealed that primary and recurrent TETs shared certain common molecular mechanisms, including a spliceosome module consisting of RNA splicing and RNA processing, but the recurrent TET was specifically related to the ribosome pathway. Applying the shortest path inference to the collected seed gene module identified that the ribonucleoprotein hnRNPA2B1 probably serves as a potential target for recurrent TET therapy. The drug repositioning combined molecular dynamics simulations suggested that the compound ergotamine could potentially act as a repurposing drug to treat recurrent TETs by targeting hnRNPA2B1. Our study demonstrates the value of integrative network analysis to understand proteotype robustness and its relationships with genotype, and provides hits for further research on cancer therapeutics.

Leveraging Protein Dynamics to Identify Functional Phosphorylation Sites using Deep Learning Models.

Accurate prediction of post-translational modifications (PTMs) is of great significance in understanding cellular processes, by modulating protein structure and dynamics. Nowadays, with the rapid growth of protein data at different "omics" levels, machine learning models largely enriched the prediction of PTMs. However, most machine learning models only rely on protein sequence and little structural information. The lack of the systematic dynamics analysis underlying PTMs largely limits the PTM functional predictions. In this research, we present two dynamics-centric deep learning models, namely, cDL-PAU and cDL-FuncPhos, by incorporating sequence, structure, and dynamics-based features to elucidate the molecular basis and underlying functional landscape of PTMs. cDL-PAU achieved satisfactory area under the curve (AUC) scores of 0.804-0.888 for predicting phosphorylation, acetylation, and ubiquitination (PAU) sites, while cDL-FuncPhos achieved an AUC value of 0.771 for predicting functional phosphorylation (FuncPhos) sites, displaying reliable improvements. Through a feature selection, the dynamics-based coupling and commute ability show large contributions in discovering PAU sites and FuncPhos sites, suggesting the allosteric propensity for important PTMs. The application of cDL-FuncPhos in three oncoproteins not only corroborates its strong performance in FuncPhos prioritization but also gains insight into the physical basis for the functions. The source code and data set of cDL-PAU and cDL-FuncPhos are available at https://github.com/ComputeSuda/PTM_ML.

Simvastatin Inhibits NLRP3 Inflammasome Activation and Ameliorates Lung Injury in Hyperoxia-Induced Bronchopulmonary Dysplasia via the KLF2-Mediated Mechanism.

Bronchopulmonary dysplasia (BPD) is a chronic lung disease commonly found in premature infants. Excessive inflammation and oxidative stress contribute to BPD occurrence and development. Simvastatin, as an inhibitor of HMG-CoA reductase, has been reported to have antioxidative and anti-inflammatory effects. However, its effect and possible mechanisms in hyperoxia-induced lung injury are rarely reported. In this study, in vivo and in vitro experiments were conducted to investigate whether simvastatin could ameliorate hyperoxia-induced lung injury and explore its potential mechanism. For the in vivo study, simvastatin could improve alveolar development after hyperoxic lung injury and reduce hyperoxic stress and inflammation. The in vitro study revealed that simvastatin can reduce inflammation in A549 cells after high-oxygen exposure. Simvastatin suppressed NLRP3 inflammasome activation and played anti-inflammatory and antioxidant roles by increasing KLF2 (Krüppel-like factor 2) expression. In vitro experiments also revealed that these effects of simvastatin were partially reversed by KLF2 shRNA, indicating that KLF2 was involved in simvastatin effects. In summary, our findings indicate that simvastatin could downregulate NLRP3 inflammasome activation and attenuate lung injury in hyperoxia-induced bronchopulmonary dysplasia via KLF2-mediated mechanism.

Protein Interaction Network Reconstruction with a Structural Gated Attention Deep Model by Incorporating Network Structure Information.

Protein-protein interactions (PPIs) provide a physical basis of molecular communications for a wide range of biological processes in living cells. Establishing the PPI network has become a fundamental but essential task for a better understanding of biological events and disease pathogenesis. Although many machine learning algorithms have been employed to predict PPIs, with only protein sequence information as the training features, these models suffer from low robustness and prediction accuracy. In this study, a new deep-learning-based framework named the Structural Gated Attention Deep (SGAD) model was proposed to improve the performance of PPI network reconstruction (PINR). The improved predictive performances were achieved by augmenting multiple protein sequence descriptors, the topological features and information flow of the PPI network, which were further implemented with a gating mechanism to improve its robustness to noise. On 11 independent test data sets and one combined data set, SGAD yielded area under the curve values of approximately 0.83-0.93, outperforming other models. Furthermore, the SGAD ensemble can learn more characteristics information on protein pairs through a two-layer neural network, serving as a powerful tool in the exploration of PPI biological space.

Dynamics of Post-Translational Modification Inspires Drug Design in the Kinase Family.

Post-translational modification (PTM) on protein plays important roles in the regulation of cellular function and disease pathogenesis. The systematic analysis of PTM dynamics presents great opportunities to enlarge the target space by PTM allosteric regulation. Here, we presented a framework by integrating the sequence, structural topology, and particular dynamics features to characterize the functional context and druggabilities of PTMs in the well-known kinase family. The machine learning models with these biophysical features could successfully predict PTMs. On the other hand, PTMs were identified to be significantly enriched in the reported allosteric pockets and the allosteric potential of PTM pockets were thus proposed through these biophysical features. In the end, the covalent inhibitor DC-Srci-6668 targeting the PTM pocket in c-Src kinase was identified, which inhibited the phosphorylation and locked c-Src in the inactive state. Our findings represent a crucial step toward PTM-inspired drug design in the kinase family.

Insights into Conformational Dynamics and Allostery in DNMT1-H3Ub/USP7 Interactions.

DNA methyltransferases (DNMTs) including DNMT1 are a conserved family of cytosine methylases that play crucial roles in epigenetic regulation. The versatile functions of DNMT1 rely on allosteric networks between its different interacting partners, emerging as novel therapeutic targets. In this work, based on the modeling structures of DNMT1-ubiquitylated H3 (H3Ub)/ubiquitin specific peptidase 7 (USP7) complexes, we have used a combination of elastic network models, molecular dynamics simulations, structural residue perturbation, network modeling, and pocket pathway analysis to examine their molecular mechanisms of allosteric regulation. The comparative intrinsic and conformational dynamics analysis of three DNMT1 systems has highlighted the pivotal role of the RFTS domain as the dynamics hub in both intra- and inter-molecular interactions. The site perturbation and network modeling approaches have revealed the different and more complex allosteric interaction landscape in both DNMT1 complexes, involving the events caused by mutational hotspots and post-translation modification sites through protein-protein interactions (PPIs). Furthermore, communication pathway analysis and pocket detection have provided new mechanistic insights into molecular mechanisms underlying quaternary structures of DNMT1 complexes, suggesting potential targeting pockets for PPI-based allosteric drug design.

Drug design targeting active posttranslational modification protein isoforms.

Modern drug design aims to discover novel lead compounds with attractable chemical profiles to enable further exploration of the intersection of chemical space and biological space. Identification of small molecules with good ligand efficiency, high activity, and selectivity is crucial toward developing effective and safe drugs. However, the intersection is one of the most challenging tasks in the pharmaceutical industry, as chemical space is almost infinity and continuous, whereas the biological space is very limited and discrete. This bottleneck potentially limits the discovery of molecules with desirable properties for lead optimization. Herein, we present a new direction leveraging posttranslational modification (PTM) protein isoforms target space to inspire drug design termed as "Post-translational Modification Inspired Drug Design (PTMI-DD)." PTMI-DD aims to extend the intersections of chemical space and biological space. We further rationalized and highlighted the importance of PTM protein isoforms and their roles in various diseases and biological functions. We then laid out a few directions to elaborate the PTMI-DD in drug design including discovering covalent binding inhibitors mimicking PTMs, targeting PTM protein isoforms with distinctive binding sites from that of wild-type counterpart, targeting protein-protein interactions involving PTMs, and hijacking protein degeneration by ubiquitination for PTM protein isoforms. These directions will lead to a significant expansion of the biological space and/or increase the tractability of compounds, primarily due to precisely targeting PTM protein isoforms or complexes which are highly relevant to biological functions. Importantly, this new avenue will further enrich the personalized treatment opportunity through precision medicine targeting PTM isoforms.

Role of protein-protein interactions in allosteric drug design for DNA methyltransferases.

DNA methyltransferases (DNMTs) not only play key roles in epigenetic gene regulation, but also serve as emerging targets for several diseases, especially for cancers. Due to the multi-domains of DNMT structures, targeting allosteric sites of protein-protein interactions (PPIs) is becoming an attractive strategy in epigenetic drug discovery. This chapter aims to review the major contemporary approaches utilized for the drug discovery based on PPIs in different dimensions, from the enumeration of allosteric mechanism to the identification of allosteric pockets. These include the construction of protein structure networks (PSNs) based on molecular dynamics (MD) simulations, performing elastic network models (ENMs) and perturbation response scanning (PRS) calculation, the sequence-based conservation and coupling analysis, and the allosteric pockets identification. Furthermore, we complement this methodology by highlighting the role of computational approaches in promising practical applications for the computer-aided drug design, with special focus on two DNMTs, namely, DNMT1 and DNMT3A.

Both intra and inter-domain interactions define the intrinsic dynamics and allosteric mechanism in DNMT1s.

DNA methyltransferase 1 (DNMT1), a large multidomain enzyme, is believed to be involved in the passive transmission of genomic methylation patterns via methylation maintenance. Yet, the molecular mechanism of interaction networks underlying DNMT1 structures, dynamics, and its biological significance has yet to be fully characterized. In this work, we used an integrated computational strategy that combined coarse-grained and atomistic simulations with coevolution information and network modeling of the residue interactions for the systematic investigation of allosteric dynamics in DNMT1. The elastic network modeling has proposed that the high plasticity of RFTS has strengthened the correlated behaviors of DNMT1 structures through the hinge sites located at the RFTS-CD interface, which mediate the collective motions between domains. The perturbation response scanning (PRS) analysis combined with the enrichment analysis of disease mutations have further highlighted the allosteric potential of the RFTS domain. Furthermore, the long-range paths connect the intra-domain interactions through the TRD interface and catalytic interface, emphasizing some key inter-domain interactions as the bridges in the global allosteric regulation of DNMT1. The observed interplay between conserved intra-domain networks and dynamical plasticity encoded by inter-domain interactions provides insights into the intrinsic dynamics and functional evolution, as well as the design of allosteric modulators of DNMT1 based on the TRD interface.

Integration of network models and evolutionary analysis into high-throughput modeling of protein dynamics and allosteric regulation: theory, tools and applications.

Proteins are dynamical entities that undergo a plethora of conformational changes, accomplishing their biological functions. Molecular dynamics simulation and normal mode analysis methods have become the gold standard for studying protein dynamics, analyzing molecular mechanism and allosteric regulation of biological systems. The enormous amount of the ensemble-based experimental and computational data on protein structure and dynamics has presented a major challenge for the high-throughput modeling of protein regulation and molecular mechanisms. In parallel, bioinformatics and systems biology approaches including genomic analysis, coevolution and network-based modeling have provided an array of powerful tools that complemented and enriched biophysical insights by enabling high-throughput analysis of biological data and dissection of global molecular signatures underlying mechanisms of protein function and interactions in the cellular environment. These developments have provided a powerful interdisciplinary framework for quantifying the relationships between protein dynamics and allosteric regulation, allowing for high-throughput modeling and engineering of molecular mechanisms. Here, we review fundamental advances in protein dynamics, network theory and coevolutionary analysis that have provided foundation for rapidly growing computational tools for modeling of allosteric regulation. We discuss recent developments in these interdisciplinary areas bridging computational biophysics and network biology, focusing on promising applications in allosteric regulations, including the investigation of allosteric communication pathways, protein-DNA/RNA interactions and disease mutations in genomic medicine. We conclude by formulating and discussing future directions and potential challenges facing quantitative computational investigations of allosteric regulatory mechanisms in protein systems.

Node-Weighted Amino Acid Network Strategy for Characterization and Identification of Protein Functional Residues.

The study of functional residues (FRs) is essential for understanding protein functions and biological processes. The amino acid network (AAN) has become an emerging paradigm for studying FRs during the past decade. Current AAN models ignore the heterogeneity of nodes and treat amino acids in the AAN as the same. However, the properties of each amino acid node are of fundamental importance. We here proposed a node-weighted AAN strategy termed the node-weighted amino acid contact energy network (NACEN) to characterize and predict three types of FRs, namely, hot spots, catalytic residues, and allosteric residues. We first constructed NACENs with their nodes weighted based on structural, sequence, physicochemical, and dynamical properties of the amino acids and then characterized the FRs with the NACEN parameters. We finally built machine learning predictors to identify each type of FR. The results revealed that residues characterized with NACEN parameters are more distinguishable between FRs and non-FRs than those with unweighted network ones. With few features for classification, NACEN yields comparable performance for FR identification and provides residue level prediction for allosteric regulation. The proposed strategy can be easily implemented to other functional residue identification. An R package is also provided for NACEN construction and analysis at http://sysbio.suda.edu.cn/NACEN/index.html .

Recent Advances on the Network Models in Target-based Drug Discovery.

With the advancement of "proteomics" data and systems biology, new techniques are needed to meet the new era of drug discovery. Network theory is increasingly applied to describe complex biological systems, thus implying its essential roles in system-based drug design. In this review, we first summarized general network parameters used in describing biological systems, and then gave some recent applications of these network parameters as topological indices in drug design in terms of Protein Structure Networks (PSNs), Protein-Protein Interaction Networks (PPINs) including related structural PPINs, and Elastic Network Models (ENMs). These network models have enabled the development of new drugs relying on allosteric effects, describing anti-cancer targets, targeting hot spots and key proteins at the protein-protein interfaces and PPINs, and helped drug design by modulating conformational flexibility. Accordingly, we highlighted the integration of network models bringing new paradigms into the next-generation target-based drug discovery.

Deciphering the role of dimer interface in intrinsic dynamics and allosteric pathways underlying the functional transformation of DNMT3A.

BACKGROUND: DNMT3A, as de novo DNA methyltransferase, is essential for regulating gene expression through cellular development and differentiation. The functions of DNMT3A rely on its oligomeric states and allosteric regulations between its catalytic domain and binding partners. Despite recent resolution of autoinhibitory and active DNMT3A/3L crystal structures, the mechanism of their functional motions and interdomain allostery in regulating the activity remains to be established. METHODS: The hybrid approach, comprising Elastic Network Models coupled with information theory, Protein Structure Network, and sequence evolution analysis was employed to investigate intrinsic dynamics and allosteric properties of DNMT3A resolved in autoinhibitory and active states. RESULTS: The conformational transition between two states is characterized by global motions, and the homo-dimer displays the similar dynamic properties as tetramer, acting as the basic functional unit. The hinge residues with restricted fluctuations are clustered at the dimer interface, which are predicted to enjoy remarkably efficient signal transduction properties. The allosteric pathways through the dimer interface are achieved by a cascade of interactions predominantly involving conserved and co-evolved residues. CONCLUSIONS: Our results suggest that structural topology coupled with global motions indicates the structural origin of the functional transformation of DNMT3A. The comprehensive analysis further highlights the pivotal role of the dimer interface of DNMT3A both in defining the quaternary structure dynamics and establishing interdomain communications. GENERAL SIGNIFICANCE: Understanding the global motions of DNMT3As not only provides mechanical insights into the functions of such molecular machines, but also reveals the mediators that determine their allosteric regulations.