Prediction and design of protease enzyme specificity using a structure-aware graph convolutional network.

Site-specific proteolysis by the enzymatic cleavage of small linear sequence motifs is a key posttranslational modification involved in physiology and disease. The ability to robustly and rapidly predict protease-substrate specificity would also enable targeted proteolytic cleavage by designed proteases. Current methods for predicting protease specificity are limited to sequence pattern recognition in experimentally derived cleavage data obtained for libraries of potential substrates and generated separately for each protease variant. We reasoned that a more semantically rich and robust model of protease specificity could be developed by incorporating the energetics of molecular interactions between protease and substrates into machine learning workflows. We present Protein Graph Convolutional Network (PGCN), which develops a physically grounded, structure-based molecular interaction graph representation that describes molecular topology and interaction energetics to predict enzyme specificity. We show that PGCN accurately predicts the specificity landscapes of several variants of two model proteases. Node and edge ablation tests identified key graph elements for specificity prediction, some of which are consistent with known biochemical constraints for protease:substrate recognition. We used a pretrained PGCN model to guide the design of protease libraries for cleaving two noncanonical substrates, and found good agreement with experimental cleavage results. Importantly, the model can accurately assess designs featuring diversity at positions not present in the training data. The described methodology should enable the structure-based prediction of specificity landscapes of a wide variety of proteases and the construction of tailor-made protease editors for site-selectively and irreversibly modifying chosen target proteins.

Prediction and Design of Protease Enzyme Specificity Using a Structure-Aware Graph Convolutional Network.

Site-specific proteolysis by the enzymatic cleavage of small linear sequence motifs is a key post-translational modification involved in physiology and disease. The ability to robustly and rapidly predict protease substrate specificity would also enable targeted proteolytic cleavage - editing - of a target protein by designed proteases. Current methods for predicting protease specificity are limited to sequence pattern recognition in experimentally-derived cleavage data obtained for libraries of potential substrates and generated separately for each protease variant. We reasoned that a more semantically rich and robust model of protease specificity could be developed by incorporating the three-dimensional structure and energetics of molecular interactions between protease and substrates into machine learning workflows. We present Protein Graph Convolutional Network (PGCN), which develops a physically-grounded, structure-based molecular interaction graph representation that describes molecular topology and interaction energetics to predict enzyme specificity. We show that PGCN accurately predicts the specificity landscapes of several variants of two model proteases: the NS3/4 protease from the Hepatitis C virus (HCV) and the Tobacco Etch Virus (TEV) proteases. Node and edge ablation tests identified key graph elements for specificity prediction, some of which are consistent with known biochemical constraints for protease:substrate recognition. We used a pre-trained PGCN model to guide the design of TEV protease libraries for cleaving two non-canonical substrates, and found good agreement with experimental cleavage results. Importantly, the model can accurately assess designs featuring diversity at positions not present in the training data. The described methodology should enable the structure-based prediction of specificity landscapes of a wide variety of proteases and the construction of tailor-made protease editors for site-selectively and irreversibly modifying chosen target proteins.

Simplified quality assessment for small-molecule ligands in the Protein Data Bank.

More than 70% of the experimentally determined macromolecular structures in the Protein Data Bank (PDB) contain small-molecule ligands. Quality indicators of ∼643,000 ligands present in ∼106,000 PDB X-ray crystal structures have been analyzed. Ligand quality varies greatly with regard to goodness of fit between ligand structure and experimental data, deviations in bond lengths and angles from known chemical structures, and inappropriate interatomic clashes between the ligand and its surroundings. Based on principal component analysis, correlated quality indicators of ligand structure have been aggregated into two largely orthogonal composite indicators measuring goodness of fit to experimental data and deviation from ideal chemical structure. Ranking of the composite quality indicators across the PDB archive enabled construction of uniformly distributed composite ranking score. This score is implemented at RCSB.org to compare chemically identical ligands in distinct PDB structures with easy-to-interpret two-dimensional ligand quality plots, allowing PDB users to quickly assess ligand structure quality and select the best exemplars.

Evolution of the SARS-CoV-2 proteome in three dimensions (3D) during the first 6 months of the COVID-19 pandemic.

Understanding the molecular evolution of the SARS-CoV-2 virus as it continues to spread in communities around the globe is important for mitigation and future pandemic preparedness. Three-dimensional structures of SARS-CoV-2 proteins and those of other coronavirusess archived in the Protein Data Bank were used to analyze viral proteome evolution during the first 6 months of the COVID-19 pandemic. Analyses of spatial locations, chemical properties, and structural and energetic impacts of the observed amino acid changes in >48 000 viral isolates revealed how each one of 29 viral proteins have undergone amino acid changes. Catalytic residues in active sites and binding residues in protein-protein interfaces showed modest, but significant, numbers of substitutions, highlighting the mutational robustness of the viral proteome. Energetics calculations showed that the impact of substitutions on the thermodynamic stability of the proteome follows a universal bi-Gaussian distribution. Detailed results are presented for potential drug discovery targets and the four structural proteins that comprise the virion, highlighting substitutions with the potential to impact protein structure, enzyme activity, and protein-protein and protein-nucleic acid interfaces. Characterizing the evolution of the virus in three dimensions provides testable insights into viral protein function and should aid in structure-based drug discovery efforts as well as the prospective identification of amino acid substitutions with potential for drug resistance.

Evaluation of Teachers' Educational Technology Ability Based on Fuzzy Clustering Generalized Regression Neural Network.

The improvement of teachers' educational technology ability is one of the main methods to improve the management efficiency of colleges and universities in China, and the scientific evaluation of teachers' ability is of great significance. In view of this, this study proposes an evaluation model of teachers' educational technology ability based on the fuzzy clustering generalized regression neural network. Firstly, the comprehensive evaluation structure system of teachers' educational technology ability is constructed, and then the prediction method of teachers' ability based on fuzzy clustering algorithm is analysed. On this basis, the optimization prediction method of fuzzy clustering generalized regression neural network is proposed. Finally, the application effect of fuzzy clustering generalized regression neural network in the evaluation of teachers' educational technology ability is analysed. The results show that the evaluation system of teachers' educational technology ability proposed in this study is scientific and reasonable; fuzzy clustering generalized regression neural network model can better accurately predict the ability of teachers' educational technology and can quickly realize global optimization. According to the fitness analysis results of the fuzzy clustering generalized regression neural network model, the model converges after the 20th iteration and the fitness value remains about 1.45. Therefore, the fuzzy clustering generalized regression neural network has stronger adaptability and has been optimized to a certain extent. The average evaluation accuracy of fuzzy clustering generalized regression neural network model is 98.44%, and the evaluation results of the model are better than other algorithms. It is hoped that this study can provide some reference value for the evaluation of teachers' educational technology ability in colleges and universities in China.

Evolution of the SARS-CoV-2 proteome in three dimensions (3D) during the first six months of the COVID-19 pandemic.

Three-dimensional structures of SARS-CoV-2 and other coronaviral proteins archived in the Protein Data Bank were used to analyze viral proteome evolution during the first six months of the COVID-19 pandemic. Analyses of spatial locations, chemical properties, and structural and energetic impacts of the observed amino acid changes in >48,000 viral proteome sequences showed how each one of the 29 viral study proteins have undergone amino acid changes. Structural models computed for every unique sequence variant revealed that most substitutions map to protein surfaces and boundary layers with a minority affecting hydrophobic cores. Conservative changes were observed more frequently in cores versus boundary layers/surfaces. Active sites and protein-protein interfaces showed modest numbers of substitutions. Energetics calculations showed that the impact of substitutions on the thermodynamic stability of the proteome follows a universal bi-Gaussian distribution. Detailed results are presented for six drug discovery targets and four structural proteins comprising the virion, highlighting substitutions with the potential to impact protein structure, enzyme activity, and functional interfaces. Characterizing the evolution of the virus in three dimensions provides testable insights into viral protein function and should aid in structure-based drug discovery efforts as well as the prospective identification of amino acid substitutions with potential for drug resistance.

Correction: Qi, W., et al. (+)-Usnic Acid Induces ROS-dependent Apoptosis via Inhibition of Mitochondria Respiratory Chain Complexes and Nrf2 Expression in Lung Squamous Cell Carcinoma. Int. J. Mol. Sci. 2020, 21, 876.

The authors wish to make the following corrections to this paper [1]:[...].

(+)-Usnic Acid Induces ROS-dependent Apoptosis via Inhibition of Mitochondria Respiratory Chain Complexes and Nrf2 Expression in Lung Squamous Cell Carcinoma.

Lung squamous cell carcinoma (LUSC) has a poor prognosis, in part due to poor therapeutic response and limited therapeutic alternatives. Lichens are symbiotic organisms, producing a variety of substances with multiple biological activities. (+)-Usnic acid, an important biologically active metabolite of lichens, has been shown to have high anti-cancer activity at low doses. However, there have been no reports regarding the effect of (+)-usnic acid on LUSC cells. This study found that (+)-usnic acid reduced viability and induced apoptosis in LUSC cells by reactive oxygen species (ROS) accumulation. (+)-Usnic acid induced mitochondria-derived ROS production via inhibition of complex I and complex III of the mitochondrial respiratory chain (MRC). Interestingly, the elimination of mitochondrial ROS by Mito-TEMPOL only partially reversed the effect of (+)-usnic acid on cellular ROS production. Further study showed that (+)-usnic acid also induced ROS production via reducing Nrf2 stability through disruption of the PI3K/Akt pathway. The in vitro and in vivo xenograft studies showed that combined treatment of (+)-usnic acid and paclitaxel synergistically suppressed LUSC cells. In conclusion, this study indicates that (+)-usnic acid induces apoptosis of LUSC cells through ROS accumulation, probably via disrupting the mitochondrial respiratory chain (MRC) and the PI3K/Akt/Nrf2 pathway. Therefore, although clinical use of (+)-usnic acid will be limited due to toxicity issues, derivatives thereof may turn out as promising anticancer candidates for adjuvant treatment of LUSC.

Sesamin suppresses NSCLC cell proliferation and induces apoptosis via Akt/p53 pathway.

Non-small cell lung cancer (NSCLC) is the most common type of lung cancer with a disappointing prognosis. The aim of this study was to investigate the anticancer effect of sesamin and the underlying mechanism. The MTT assay was used to detect the proliferation of NSCLC cells. The cell cycle and apoptosis were analyzed by flow cytometry. The protein levels of Akt, p-Akt (Ser473), p53, cyclin D1, CDK2, MDM2, p-MDM2 (Ser166) were detected by western blotting. The expression of p-Akt (Ser473), p53 and Ki67 in vivo was analyzed by IHC. Histopathologic analyses of major organs (heart, liver, spleen, lung and kidney) were performed by H&E staining. The results show that sesamin suppressed cell proliferation and induced apoptosis of NSCLC cells (A549 and H1792) in a dose-dependent manner. Treatment with sesamin caused cell cycle arrest at G1 phase and inhibited cyclin D1 and CDK2 expression. In addition, sesamin inhibited Akt activity and upregulated p53 expression both in vivo and in vitro. When Akt and p53 were suppressed by LY294002 and PFTα, respectively, sesamin exerted no additional effects. The in vivo results mostly matched the in vitro findings. Specifically, sesamin exerted little damage to major organs. Taken together, this study demonstrates that sesamin suppresses NSCLC cell proliferation by induction of G1 phase cell cycle arrest and apoptosis via Akt/p53 pathway. Therefore, sesamin may be a promising adjuvant treatment for NSCLC therapy.