DeepSite: protein-binding site predictor using 3D-convolutional neural networks

Structural and functional prediction, evaluation, and validation in the post-sequencing era

The surge of genome sequencing data has underlined substantial genetic variants of uncertain significance (VUS). The decryption of VUS discovered by sequencing poses a major challenge in the post-sequencing era. Although experimental assays have progressed in classifying VUS, only a tiny fraction of the human genes have been explored experimentally. Thus, it is urgently needed to generate state-of-the-art functional predictors of VUS in silico. Artificial intelligence (AI) is an invaluable tool to assist in the identification of VUS with high efficiency and accuracy. An increasing number of studies indicate that AI has brought an exciting acceleration in the interpretation of VUS, and our group has already used AI to develop protein structure-based prediction models. In this review, we provide an overview of the previous research on AI-based prediction of missense variants, and elucidate the challenges and opportunities for protein structure-based variant prediction in the post-sequencing era.

Databases of ligand-binding pockets and protein-ligand interactions

Many research groups and institutions have created a variety of databases curating experimental and predicted data related to protein-ligand binding. The landscape of available databases is dynamic, with new databases emerging and established databases becoming defunct. Here, we review the current state of databases that contain binding pockets and protein-ligand binding interactions. We have compiled a list of such databases, fifty-three of which are currently available for use. We discuss variation in how binding pockets are defined and summarize pocket-finding methods. We organize the fifty-three databases into subgroups based on goals and contents, and describe standard use cases. We also illustrate that pockets within the same protein are characterized differently across different databases. Finally, we assess critical issues of sustainability, accessibility and redundancy.

Language models can identify enzymatic binding sites in protein sequences

Recent advances in language modeling have had a tremendous impact on how we handle sequential data in science. Language architectures have emerged as a hotbed of innovation and creativity in natural language processing over the last decade, and have since gained prominence in modeling proteins and chemical processes, elucidating structural relationships from textual/sequential data. Surprisingly, some of these relationships refer to three-dimensional structural features, raising important questions on the dimensionality of the information encoded within sequential data. Here, we demonstrate that the unsupervised use of a language model architecture to a language representation of bio-catalyzed chemical reactions can capture the signal at the base of the substrate-binding site atomic interactions. This allows us to identify the three-dimensional binding site position in unknown protein sequences. The language representation comprises a reaction-simplified molecular-input line-entry system (SMILES) for substrate and products, and amino acid sequence information for the enzyme. This approach can recover, with no supervision, 52.13% of the binding site when considering co-crystallized substrate-enzyme structures as ground truth, vastly outperforming other attention-based models.

Engineering hyperthermophilic pullulanase to efficiently utilize corn starch for production of maltooligosaccharides and glucose

Pullulanase is a starch-debranching enzyme that hydrolyzes side chain of starch, oligosaccharides and pullulan. Nevertheless, the limited activities of pullulanases constrain their practical application. Herein, the hyperthermophilic type II pullulanase from Pyrococcus yayanosii CH1 (PulPY2) was evolved by synergistically engineering the substrate-binding pocket and active-site lids. The resulting mutant PulPY2-M2 exhibited 5-fold improvement in catalytic efficiency (kcat/Km) compared to that of PulPY2. PulPY2-M2 was utilized to develop a one-pot reaction system for efficient production of maltooligosaccharides. The maltooligosaccharides conversion rate of PulPY2-M2 reached 96.1%, which was increased by 5.4% compared to that of PulPY2. Furthermore, when employed for glucose production, the glucose productivity of PulPY2-M2 was 25.4% and 43.5% higher than that of PulPY2 and the traditional method, respectively. These significant improvements in maltooligosaccharides and glucose production and the efficient utilization of corn starch demonstrated the potential of the engineered PulPY2-M2 in starch sugar industry.

PUResNetV2.0: a deep learning model leveraging sparse representation for improved ligand binding site prediction

Accurate ligand binding site prediction (LBSP) within proteins is essential for drug discovery. We developed ProteinUNetResNetV2.0 (PUResNetV2.0), leveraging sparse representation of protein structures to improve LBSP accuracy. Our training dataset included protein complexes from 4729 protein families. Evaluations on benchmark datasets showed that PUResNetV2.0 achieved an 85.4% Distance Center Atom (DCA) success rate and a 74.7% F1 Score on the Holo801 dataset, outperforming existing methods. However, its performance in specific cases, such as RNA, DNA, peptide-like ligand, and ion binding site prediction, was limited due to constraints in our training data. Our findings underscore the potential of sparse representation in LBSP, especially for oligomeric structures, suggesting PUResNetV2.0 as a promising tool for computational drug discovery.

A comprehensive survey on protein-ligand binding site prediction

Protein-ligand binding site prediction is critical for protein function annotation and drug discovery. Biological experiments are time-consuming and require significant equipment, materials, and labor resources. Developing accurate and efficient computational methods for protein-ligand interaction prediction is essential. Here, we summarize the key challenges associated with ligand binding site (LBS) prediction and introduce recently published methods from their input features, computational algorithms, and ligand types. Furthermore, we investigate the specificity of allosteric site identification as a particular LBS type. Finally, we discuss the prospective directions for machine learning-based LBS prediction in the near future.

From GPUs to AI and quantum: three waves of acceleration in bioinformatics

The enormous growth in the amount of data generated by the life sciences is continuously shifting the field from model-driven science towards data-driven science. The need for efficient processing has led to the adoption of massively parallel accelerators such as graphics processing units (GPUs). Consequently, the development of bioinformatics methods nowadays often heavily depends on the effective use of these powerful technologies. Furthermore, progress in computational techniques and architectures continues to be highly dynamic, involving novel deep neural network models and artificial intelligence (AI) accelerators, and potentially quantum processing units in the future. These are expected to be disruptive for the life sciences as a whole and for drug discovery in particular. Here, we identify three waves of acceleration and their applications in a bioinformatics context: (i) GPU computing, (ii) AI and (iii) next-generation quantum computers.

Integrating genome-wide CRISPR screens and in silico drug profiling for targeted antidote development

Numerous toxins threaten humans, but specific antidotes are unavailable for most of them. Although CRISPR screening has aided the discovery of the mechanisms of some toxins, developing targeted antidotes remains a significant challenge. Recently, we established a systematic framework to develop antidotes by combining the identification of novel drug targets by using a genome-wide CRISPR screen with a virtual screen of drugs approved by the US Food and Drug Administration. This approach allows for a comprehensive understanding of toxin mechanisms at the whole-genome level and facilitates the identification of promising antidote drugs targeting specific molecules. Here, we present step-by-step instructions for executing genome-scale CRISPR-Cas9 knockout screens of toxins in HAP1 cells. We also provide detailed guidance for conducting an in silico drug screen and an in vivo drug validation. By using this protocol, it takes ~4 weeks to perform the genome-scale screen, 4 weeks for sequencing and data analysis, 4 weeks to validate candidate genes, 1 week for the virtual screen and 2 weeks for in vitro drug validation. This framework has the potential to accelerate the development of antidotes for a wide range of toxins and can rapidly identify promising drug candidates that are already known to be safe and effective. This could lead to the development of new antidotes much more quickly than traditional methods, protecting lives from diverse toxins and advancing human health.

Annona squamosa Fruit Extract Ameliorates Lead Acetate-Induced Testicular Injury by Modulating JAK-1/STAT-3/SOCS-1 Signaling in Male Rats

Lead (Pb) is a common pollutant that is not biodegradable and gravely endangers the environment and human health. Annona squamosa fruit has a wide range of medicinal uses owing to its phytochemical constituents. This study evaluated the effect of treatment with A. squamosa fruit extract (ASFE) on testicular toxicity induced in male rats by lead acetate. The metal-chelating capacity and phytochemical composition of ASFE were determined. The LD50 of ASFE was evaluated by probit analysis. Molecular docking simulations were performed using Auto Dock Vina. Forty male Sprague Dawley rats were equally divided into the following groups: Gp1, a negative control group; Gp2, given ASFE (350 mg/kg body weight (b. wt.)) (1/10 of LD50); Gp3, given lead acetate (PbAc) solution (100 mg/kg b. wt.); and Gp4, given PbAc as in Gp3 and ASFE as in Gp2. All treatments were given by oro-gastric intubation daily for 30 days. Body weight changes, spermatological parameters, reproductive hormone levels, oxidative stress parameters, and inflammatory biomarkers were evaluated, and molecular and histopathological investigations were performed. The results showed that ASFE had promising metal-chelating activity and phytochemical composition. The LD50 of ASFE was 3500 mg/kg b. wt. The docking analysis showed that quercetin demonstrated a high binding affinity for JAK-1 and STAT-3 proteins, and this could make it a more promising candidate for targeting the JAK-1/STAT-3 pathway than others. The rats given lead acetate had defective testicular tissues, with altered molecular, biochemical, and histological features, as well as impaired spermatological characteristics. Treatment with ASFE led to a significant mitigation of these dysfunctions and modulated the JAK-1/STAT-3/SOCS-1 axis in the rats.

Repurposing DrugBank compounds as potential Plasmodium falciparum class 1a aminoacyl tRNA synthetase multi-stage pan-inhibitors with a specific focus on mitomycin

Plasmodium falciparum aminoacyl tRNA synthetases (PfaaRSs) are potent antimalarial targets essential for proteome fidelity and overall parasite survival in every stage of the parasite's life cycle. So far, some of these proteins have been singly targeted yielding inhibitor compounds that have been limited by incidences of resistance which can be overcome via pan-inhibition strategies. Hence, herein, for the first time, we report the identification and in vitro antiplasmodial validation of Mitomycin (MMC) as a probable pan-inhibitor of class 1a (arginyl(A)-, cysteinyl(C), isoleucyl(I)-, leucyl(L), methionyl(M), and valyl(V)-) PfaaRSs which hypothetically may underlie its previously reported activity on the ribosomal RNA to inhibit protein translation and biosynthesis. We combined multiple in silico structure-based discovery strategies that first helped identify functional and druggable sites that were preferentially targeted by the compound in each of the plasmodial proteins: Ins1-Ins2 domain in Pf-ARS; anticodon binding domain in Pf-CRS; CP1-editing domain in Pf-IRS and Pf-MRS; C-terminal domain in Pf-LRS; and CP-core region in Pf-VRS. Molecular dynamics studies further revealed that MMC allosterically induced changes in the global structures of each protein. Likewise, prominent structural perturbations were caused by the compound across the functional domains of the proteins. More so, MMC induced systematic alterations in the binding of the catalytic nucleotide and amino acid substrates which culminated in the loss of key interactions with key active site residues and ultimate reduction in the nucleotide-binding affinities across all proteins, as deduced from the binding energy calculations. These altogether confirmed that MMC uniformly disrupted the structure of the target proteins and essential substrates. Further, MMC demonstrated IC50 < 5 μM against the Dd2 and 3D7 strains of parasite making it a good starting point for malarial drug development. We believe that findings from our study will be important in the current search for highly effective multi-stage antimalarial drugs.

Advances in machine intelligence-driven virtual screening approaches for big-data

Virtual screening (VS) is an integral and ever-evolving domain of drug discovery framework. The VS is traditionally classified into ligand-based (LB) and structure-based (SB) approaches. Machine intelligence or artificial intelligence has wide applications in the drug discovery domain to reduce time and resource consumption. In combination with machine intelligence algorithms, VS has emerged into revolutionarily progressive technology that learns within robust decision orders for data curation and hit molecule screening from large VS libraries in minutes or hours. The exponential growth of chemical and biological data has evolved as "big-data" in the public domain demands modern and advanced machine intelligence-driven VS approaches to screen hit molecules from ultra-large VS libraries. VS has evolved from an individual approach (LB and SB) to integrated LB and SB techniques to explore various ligand and target protein aspects for the enhanced rate of appropriate hit molecule prediction. Current trends demand advanced and intelligent solutions to handle enormous data in drug discovery domain for screening and optimizing hits or lead with fewer or no false positive hits. Following the big-data drift and tremendous growth in computational architecture, we presented this review. Here, the article categorized and emphasized individual VS techniques, detailed literature presented for machine learning implementation, modern machine intelligence approaches, and limitations and deliberated the future prospects.

Binding Sites of Bicarbonate in Phosphoenolpyruvate Carboxylase

Phosphoenolpyruvate carboxylase (PEPC) is used in plant metabolism for fruit maturation or seed development as well as in the C4 and crassulacean acid metabolism (CAM) mechanisms in photosynthesis, where it is used for the capture of hydrated CO2 (bicarbonate). To find the yet unknown binding site of bicarbonate in this enzyme, we have first identified putative binding sites with nonequilibrium molecular dynamics simulations and then ranked these sites with alchemical free energy calculations with corrections of computational artifacts. Fourteen pockets where bicarbonate could bind were identified, with three having realistic binding free energies with differences with the experimental value below 1 kcal/mol. One of these pockets is found far from the active site at 14 Å and predicted to be an allosteric binding site. In the two other binding sites, bicarbonate is in direct interaction with the magnesium ion; neither sequence alignment nor the study of mutant K606N allowed to discriminate between these two pockets, and both are good candidates as the binding site of bicarbonate in phosphoenolpyruvate carboxylase.

Genome-scale annotation of protein binding sites via language model and geometric deep learning

Revealing protein binding sites with other molecules, such as nucleic acids, peptides, or small ligands, sheds light on disease mechanism elucidation and novel drug design. With the explosive growth of proteins in sequence databases, how to accurately and efficiently identify these binding sites from sequences becomes essential. However, current methods mostly rely on expensive multiple sequence alignments or experimental protein structures, limiting their genome-scale applications. Besides, these methods haven't fully explored the geometry of the protein structures. Here, we propose GPSite, a multi-task network for simultaneously predicting binding residues of DNA, RNA, peptide, protein, ATP, HEM, and metal ions on proteins. GPSite was trained on informative sequence embeddings and predicted structures from protein language models, while comprehensively extracting residual and relational geometric contexts in an end-to-end manner. Experiments demonstrate that GPSite substantially surpasses state-of-the-art sequence-based and structure-based approaches on various benchmark datasets, even when the structures are not well-predicted. The low computational cost of GPSite enables rapid genome-scale binding residue annotations for over 568,000 sequences, providing opportunities to unveil unexplored associations of binding sites with molecular functions, biological processes, and genetic variants. The GPSite webserver and annotation database can be freely accessed at https://bio-web1.nscc-gz.cn/app/GPSite.

Identification of a novel FOXO3 agonist that protects against alcohol induced liver injury

Alcohol-related liver disease (ALD) is a global healthcare concern which caused by excessive alcohol consumption with limited treatment options. The pathogenesis of ALD is complex and involves in hepatocyte damage, hepatic inflammation, increased gut permeability and microbiome dysbiosis. FOXO3 is a well-recognized transcription factor which associated with longevity via promoting antioxidant stress response, preventing senescence and cell death, and inhibiting inflammation. We and many others have reported that FOXO3-/- mice develop more severe liver injury in response to alcohol. In the present study, we aimed to develop compounds that activate FOXO3 and further investigate their effects in alcohol induced liver injury. Through virtual screening, we discovered series of small molecular compounds that showed high affinity to FOXO3. We confirmed effects of compounds on FOXO3 target gene expression, as well as antioxidant and anti-apoptotic effects in vitro. Subsequently we evaluated the protective efficacy of compounds in alcohol induced liver injury in vivo. As a result, the leading compound we identified, 214991, activated downstream target genes expression of FOXO3, inhibited intracellular ROS accumulation and cell apoptosis induced by H2O2 and sorafenib. By using Lieber-DeCarli alcohol feeding mouse model, 214991 showed protective effects against alcohol-induced liver inflammation, macrophage and neutrophil infiltration, and steatosis. These findings not only reinforce the potential of FOXO3 as a valuable target for therapeutic intervention of ALD, but also suggested that compound 214991 as a promising candidate for the development of innovative therapeutic strategies of ALD.

Small molecule drug discovery for glioblastoma treatment based on bioinformatics and cheminformatics approaches

Background: Glioblastoma (GBM) is a common and highly aggressive brain tumor with a poor prognosis for patients. It is urgently needed to identify potential small molecule drugs that specifically target key genes associated with GBM development and prognosis. Methods: Differentially expressed genes (DEGs) between GBM and normal tissues were obtained by data mining the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. Gene function annotation was performed to investigate the potential functions of the DEGs. A protein-protein interaction (PPI) network was constructed to explore hub genes associated with GBM. Bioinformatics analysis was used to screen the potential therapeutic and prognostic genes. Finally, potential small molecule drugs were predicted using the DGIdb database and verified using chemical informatics methods including absorption, distribution, metabolism, excretion, toxicity (ADMET), and molecular docking studies. Results: A total of 429 DEGs were identified, of which 19 hub genes were obtained through PPI analysis. The hub genes were confirmed as potential therapeutic targets by functional enrichment and mRNA expression. Survival analysis and protein expression confirmed centromere protein A (CENPA) as a prognostic target in GBM. Four small molecule drugs were predicted for the treatment of GBM. Conclusion: Our study suggests some promising potential therapeutic targets and small molecule drugs for the treatment of GBM, providing new ideas for further research and targeted drug development.

Graph Attention Site Prediction (GrASP): Identifying Druggable Binding Sites Using Graph Neural Networks with Attention

Identifying and discovering druggable protein binding sites is an important early step in computer-aided drug discovery, but it remains a difficult task where most campaigns rely on a priori knowledge of binding sites from experiments. Here, we present a binding site prediction method called Graph Attention Site Prediction (GrASP) and re-evaluate assumptions in nearly every step in the site prediction workflow from data set preparation to model evaluation. GrASP is able to achieve state-of-the-art performance at recovering binding sites in PDB structures while maintaining a high degree of precision which will minimize wasted computation in downstream tasks such as docking and free energy perturbation.

Characterization and structural identification of a family 16 carbohydrate-binding module (CBM): First structural insights into porphyran-binding CBM

Porphyran is a favorable functional polysaccharide widely distributed in Porphyra. It displays a linear structure majorly constituted by alternating 1,4-linked α-l-galactopyranose-6-sulfate (L6S) and 1,3-linked β-d-galactopyranose (G) units. Carbohydrate-binding modules (CBMs) are desired tools for the investigation and application of polysaccharides, including in situ visualization, on site and specific assay, and functionalization of biomaterials. However, only one porphyran-binding CBM has been hitherto reported, and its structural knowledge is lacking. Herein, a novel CBM16 family domain from a marine bacterium Aquimarina sp. BL5 was discovered and expressed. The recombinant protein AmCBM16 exhibited the desired specificity for porphyran. Bio-layer interferometry assay revealed that the protein binds to porphyran tetrasaccharide (L6S-G)2 with an association constant of 1.3 × 103 M-1. The structure of AmCBM16 was resolved by the X-ray crystallography, which displays a β-sandwich fold with two antiparallel β-sheets constituted by 10 β-strands. Site-directed mutagenesis analysis demonstrated that the residues Gly-30, Trp-31, Lys-88, Lys-123, Phe-125, and Phe-127 play dominant roles in AmCBM16 binding. This study provides the first structural insights into porphyran-binding CBM.

Contrastive pre-training and 3D convolution neural network for RNA and small molecule binding affinity prediction

MOTIVATION

The diverse structures and functions inherent in RNAs present a wealth of potential drug targets. Some small molecules are anticipated to serve as leading compounds, providing guidance for the development of novel RNA-targeted therapeutics. Consequently, the determination of RNA-small molecule binding affinity is a critical undertaking in the landscape of RNA-targeted drug discovery and development. Nevertheless, to date, only one computational method for RNA-small molecule binding affinity prediction has been proposed. The prediction of RNA-small molecule binding affinity remains a significant challenge. The development of a computational model is deemed essential to effectively extract relevant features and predict RNA-small molecule binding affinity accurately.

RESULTS

In this study, we introduced RLaffinity, a novel deep learning model designed for the prediction of RNA-small molecule binding affinity based on 3D structures. RLaffinity integrated information from RNA pockets and small molecules, utilizing a 3D convolutional neural network (3D-CNN) coupled with a contrastive learning-based self-supervised pre-training model. To the best of our knowledge, RLaffinity was the first deep learning based method for the prediction of RNA-small molecule binding affinity. Our experimental results exhibited RLaffinity's superior performance compared to baseline methods, revealed by all metrics. The efficacy of RLaffinity underscores the capability of 3D-CNN to accurately extract both global pocket information and local neighbor nucleotide information within RNAs. Notably, the integration of a self-supervised pre-training model significantly enhanced predictive performance. Ultimately, RLaffinity was also proved as a potential tool for RNA-targeted drugs virtual screening.

AVAILABILITY AND IMPLEMENTATION

https://github.com/SaisaiSun/RLaffinity.

A comprehensive review of protein-centric predictors for biomolecular interactions: from proteins to nucleic acids and beyond

Proteins interact with diverse ligands to perform a large number of biological functions, such as gene expression and signal transduction. Accurate identification of these protein-ligand interactions is crucial to the understanding of molecular mechanisms and the development of new drugs. However, traditional biological experiments are time-consuming and expensive. With the development of high-throughput technologies, an increasing amount of protein data is available. In the past decades, many computational methods have been developed to predict protein-ligand interactions. Here, we review a comprehensive set of over 160 protein-ligand interaction predictors, which cover protein-protein, protein-nucleic acid, protein-peptide and protein-other ligands (nucleotide, heme, ion) interactions. We have carried out a comprehensive analysis of the above four types of predictors from several significant perspectives, including their inputs, feature profiles, models, availability, etc. The current methods primarily rely on protein sequences, especially utilizing evolutionary information. The significant improvement in predictions is attributed to deep learning methods. Additionally, sequence-based pretrained models and structure-based approaches are emerging as new trends.

Chromatographic Fingerprinting of Cacao Pod Husk Extracts (Theobroma cacao L.): Exploring Antibacterial, Antioxidant, and Antidiabetic Properties with In Silico Molecular Docking Analysis

Natural drugs derived from plants are becoming more popular because of their apparent biological efficacy, affordability, and safety. A byproduct of cocoa farms, cocoa pod husk (CPH), is often disregarded yet contains an abundance of phenolic chemicals that have antimicrobial and antioxidant features, which has led to intensive investigation into possible biomedical applications. In order to identify crucial functional groups and phytochemical components, we carefully examined the 80% ethanol and dichloromethane extracts of CPH using gas chromatography-mass spectrometry (GC-MS) and HPLC. The antibacterial and antioxidant properties of such extracts and their impact on cytotoxicity and α-glucosidase were explored. According to our results, the 80% ethanol and dichloromethane extracts contained 19 and 12 phytochemical components, respectively. Interestingly, at 250 µg/mL, all CPH extracts showed strong antibacterial properties that totally prevented the bacterial growth. At 66.6% and 82.7%, respectively, the ethanol and dichloromethane extracts showed impressive antioxidant and DPPH scavenging capabilities where the ethanol extract showed a substantially lower IC50 value of 35.26 µg/mL than the dichloromethane extract, which had an IC50 value of 23.88 µg/mL. Furthermore, the α-glucosidase inhibitory effect of the dichloromethane extract was found to be better, as shown by its IC50 value of 126.5 µg/mL, which was lower than that of the ethanol extract at 151.3 µg/mL. The extracts' compatibility was verified by cytotoxicity tests, which revealed no appreciable alterations in the cell lines. Additionally, novel in silico molecular docking experiments were performed on 25 discovered compounds, providing insight into their possible bioactivity. Broad-spectrum activities of extracts were confirmed by molecular docking investigations aimed at interacting with α-glucosidase proteins. Our thorough analysis makes CPH extracts seem like the excellent candidates for biomedical uses. These results provide new insights into the therapeutic potential of CPH extracts and pave the way for the development of innovative medications and natural remedies.

EquiPNAS: improved protein-nucleic acid binding site prediction using protein-language-model-informed equivariant deep graph neural networks

Protein language models (pLMs) trained on a large corpus of protein sequences have shown unprecedented scalability and broad generalizability in a wide range of predictive modeling tasks, but their power has not yet been harnessed for predicting protein-nucleic acid binding sites, critical for characterizing the interactions between proteins and nucleic acids. Here, we present EquiPNAS, a new pLM-informed E(3) equivariant deep graph neural network framework for improved protein-nucleic acid binding site prediction. By combining the strengths of pLM and symmetry-aware deep graph learning, EquiPNAS consistently outperforms the state-of-the-art methods for both protein-DNA and protein-RNA binding site prediction on multiple datasets across a diverse set of predictive modeling scenarios ranging from using experimental input to AlphaFold2 predictions. Our ablation study reveals that the pLM embeddings used in EquiPNAS are sufficiently powerful to dramatically reduce the dependence on the availability of evolutionary information without compromising on accuracy, and that the symmetry-aware nature of the E(3) equivariant graph-based neural architecture offers remarkable robustness and performance resilience. EquiPNAS is freely available at https://github.com/Bhattacharya-Lab/EquiPNAS.

Nobiletin alleviates atherosclerosis by inhibiting lipid uptake via the PPARG/CD36 pathway

BACKGROUND

Atherosclerosis (AS) is a persistent inflammatory condition triggered and exacerbated by several factors including lipid accumulation, endothelial dysfunction and macrophages infiltration. Nobiletin (NOB) has been reported to alleviate atherosclerosis; however, the underlying mechanism remains incompletely understood.

METHODS

This study involved comprehensive bioinformatic analysis, including multidatabase target prediction; GO and KEGG enrichment analyses for function and pathway exploration; DeepSite and AutoDock for drug binding site prediction; and CIBERSORT for immune cell involvement. In addition, target intervention was verified via cell scratch assays, oil red O staining, ELISA, flow cytometry, qRT‒PCR and Western blotting. In addition, by establishing a mouse model of AS, it was demonstrated that NOB attenuated lipid accumulation and the extent of atherosclerotic lesions.

RESULTS

(1) Altogether, 141 potentially targetable genes were identified through which NOB could intervene in atherosclerosis. (2) Lipid and atherosclerosis, fluid shear stress and atherosclerosis may be the dominant pathways and potential mechanisms. (3) ALB, AKT1, CASP3 and 7 other genes were identified as the top 10 target genes. (4) Six genes, including PPARG, MMP9, SRC and 3 other genes, were related to the M0 fraction. (5) CD36 and PPARG were upregulated in atherosclerosis samples compared to the normal control. (6) By inhibiting lipid uptake in RAW264.7 cells, NOB prevents the formation of foam cell. (7) In RAW264.7 cells, the inhibitory effect of oxidized low-density lipoprotein on foam cells formation and lipid accumulation was closely associated with the PPARG signaling pathway. (8) In vivo validation showed that NOB significantly attenuated intra-arterial lipid accumulation and macrophage infiltration and reduced CD36 expression.

CONCLUSIONS

Nobiletin alleviates atherosclerosis by inhibiting lipid uptake via the PPARG/CD36 pathway.

Prediction of key amino acids of Salmonella phage endolysin LysST-3 and detection of its mutants' activity

Salmonella Typhimurium, a zoonotic pathogen, causes systemic and localized infection. The emergence of drug-resistant S. Typhimurium has increased; treating bacterial infections remains challenging. Phage endolysins derived from phages have a broader spectrum of bacteriolysis and better bacteriolytic activity than phages, and are less likely to induce drug resistance than antibiotics. LysST-3, the endolysin of Salmonella phage ST-3, was chosen in our study for its high lytic activity, broad cleavage spectrum, excellent bioactivity, and moderate safety profile. LysST-3 is a promising antimicrobial agent for inhibiting the development of drug resistance in Salmonella. The aim of this study is to investigate the molecular characteristics of LysST-3 through the prediction of key amino acid sites of LysST-3 and detection of its mutants' activity. We investigated its lytic effect on Salmonella and identified its key amino acid sites of interaction with substrate. LysST-3 may be a Ca2+, Mg2+ - dependent metalloenzyme. Its concave structure of the bottom "gripper" was found to be an important part of its amino acid active site. We identified its key sites (29P, 30T, 86D, 88 L, and 89 V) for substrate binding and activity using amino acid-targeted mutagenesis. Alterations in these sites did not affect protein secondary structure, but led to a significant reduction in the cleavage activity of the mutant proteins. Our study provides a basis for phage endolysin modification to target drug-resistant bacteria. Identifying the key amino acid site of the endolysin LysST-3 provides theoretical support for the functional modification of the endolysin and the development of subsequent effective therapeutic solutions.

Classifying Protein-Protein Binding Affinity with Free-Energy Calculations and Machine Learning Approaches

Understanding the intricate phenomenon of neuronal wiring in the brain is of great interest in neuroscience. In the fruit fly, Drosophila melanogaster, the Dpr-DIP interactome has been identified to play an important role in this process. However, experimental data suggest that a merely limited subset of complexes, essentially 57 out of a total of 231, exhibit strong binding affinity. In this work, we sought to identify the residue-level molecular basis underlying the difference in binding affinity using a state-of-the-art methodology consisting of standard binding free-energy calculations with a geometrical route and machine learning (ML) techniques. We determined the binding affinity for two complexes using statistical mechanics simulations, achieving an excellent reproduction of the experimental data. Moreover, we predicted the binding free energy for two additional low-affinity complexes, devoid of experimental estimation, while simultaneously identifying key residues for the binding. Furthermore, through the use of ML algorithms, linear discriminant analysis, and random forest, we achieved remarkable accuracy, as high as 0.99, in discerning between strong (cognate) and weak (noncognate) binders. The presented ML approach encompasses easily transferable input features, enabling its broad application to any interactome while facilitating the identification of pivotal residues critical for binding interactions. The predictive power of the generated model was probed on similar protein families from 13 diverse species. Our ML model exhibited commendable performance on these additional data sets, showcasing its reliability and robustness across the species barrier.

The Odorant Receptor Recognizing Camphor in a Camphor Tree Specialist Orthaga achatina (Lepidoptera: Pyralidae)

Camphor has been used as an effective repellent and pesticide to stored products for a long history, but Orthaga achatina (Lepidoptera: Pyralidae) has evolved to specifically feed on the camphor tree Cinnamomum camphora. However, the behavioral response of O. achatina to camphor and the molecular basis of camphor perception are totally unknown. Here, we demonstrated that both male and female adults were behaviorally attracted to camphor, suggesting the adaptation of O. achatina to and utilization of camphor as a signal of C. camphora. Second, in 40 O. achatina OR genes obtained by analyzing antenna transcriptomes, only OachOR16/Orco significantly responded to camphor in the Xenopus oocyte system. Finally, by molecular docking analysis and site-directed mutagenesis, the Ser209 residue is confirmed to be essential for binding of the oachOR16 with camphor. This study not only reveals the camphor-based host plant choice and olfactory mechanisms of O. achatina but also provides a molecular target for screening more potential insect repellents.

High-efficiency removal of pyrethroids using a redesigned odorant binding protein

Pyrethroids are ubiquitously present in environmental media and threaten both the ecosystem and human health. To explore effective ways to remove pyrethroids from the environment, an odorant binding protein (OBP) with affinity for various pyrethroids was investigated. Initially, the target OBP, Spodoptera littoralis pheromone binding protein 1 (SlitPBP1), underwent redesign to enhance its affinity for pyrethroids. The modified SlitPBP1E97ND106E demonstrated a substantially increased affinity for deltamethrin (DeltaM), with a dissociation constant of 0.77 ± 0.17 μM. The affinity of SlitPBP1E97ND106E for other pyrethroids also increased to varying extents. Consequently, SlitPBP1E97ND106E displayed a markedly enhanced capability to adsorb and remove pyrethroids. When exposed to free SlitPBP1E97ND106E in solution, the reduction in DeltaM surged from 16.78 ± 0.32% to 97.51 ± 0.56%. SlitPBP1E97ND106E was immobilized by coupling the protein to Ni2+-NTA agarose resin. Liquid chromatography results attested to the superior efficacy of immobilized SlitPBP1E97ND106E in removing pyrethroids, especially DeltaM. No significant differences in pyrethroid removal were detected across various water samples. Our findings introduce a potent tool for pyrethroid removal. A wider range of OBPs can similarly be optimized and applied to remove organic pollutants, including but not limited to pesticides.

The evaluations of the inhibition of orlistat on Clostridium perfringens sialidase (NanI) activity by in vitro and in silico approaches

Clostridium perfringens (C. perfringens) is a bacterium that causes serious problems in humans and animals such as food poisoning, gas gangrene and infections. C. perfringens has three sialidases (NanH, NanI, NanJ) and inhibition of NanI constitutes an approach in the treatment of C. perfringens since NanI provides the carbohydrate source necessary for the growth of bacteria. In our study, the inhibition effect of some drugs belonging to different drug groups on NanI activity was investigated. Among these drugs, orlistat (0.21±0.05 μM) was determined to have a lower IC50 value than the positive control quercetin (15.58±1.59 μM). It was determined in vitro by spectrofluorometric method. Additionally, NanI molecular docking studies with orlistatand quercetin were performed using iGemdock, DockThor and SwissDock. Orlistat (-93.93, -8.649 and -10.03 kcal/mol, respectively) was found to have a higher binding affinity than quercetin (-92.68, -7.491 and -8.70 kcal/mol, respectively), and the results were in line with in vitro studies. The results may suggest that orlistat is a molecule with drug potential for C. perfringens because it inhibits the drug target NanI, and that the inhibition efficiency can be increased by studies with orlistat derivatives.

The in vitro and in vivo antiviral effects of IGF1R inhibitors against respiratory syncytial virus infection

The insulin-like growth factor 1 receptor (IGF1R) was recognized as a pivotal receptor that facilitated the cellular entry of RSV. Small molecule inhibitors designed to target IGF1R exhibited potential as potent antiviral agents. Through virtual screening, we conducted a screening process involving small molecule compounds derived from natural products, aiming to target the IGF1R protein against respiratory syncytial virus infection. The molecular dynamics simulation analysis showed that tannic acid and daptomycin interacted with the IGF1R. The experimental results in vivo and in vitro showed that tannic acid and daptomycin had anti-RSV infection potential through reducing viral loads, inflammation, airway resistance and protecting alveolar integrity. The CC50 values of tannic acid and daptomycin were 6 nM and 0.45 μM, respectively. Novel small-molecule inhibitors targeting the IGF1R, tannic acid and daptomycin, may be effective anti-RSV therapy agents. This study may in future broaden the arsenal of therapeutics for use against RSV infection and lead to more effective care against the virus.Communicated by Ramaswamy H. Sarma.

Chicoric acid inserted in protein Z cavity exhibits higher stability and better wound healing effect under oxidative stress

Oxidative stress is one of the limiting factors that inhibit wound healing. Phytochemicals especially chicoric acid have the potential to act as an antioxidant and scavenge reactive oxygen species, thereby promoting wound healing. However, most of the phytochemicals were easy to be degraded during storage or using due to the oxidative status in wound site. Herein, we introduce a high stable protein Z that can encapsulate chicoric acid during foaming. TEM results showed that the size of protein Z-chicoric acid is in the range of nanoscale (named PZ-CA nanocomposite), and protein Z encapsulation can significantly improve the stability of chicoric acid under oxidative stress. Moreover, PZ-CA nanocomposite exhibited favorable antioxidant properties, biocompatibility, and the ability to promote cell migration in vitro. The role of PZ-CA nanocomposite in skin regeneration was explored by a mice model. Results in vivo suggest that the PZ-CA nanocomposite promotes wound healing with a faster rate as compared with a commercial spray solution, mostly through attenuating the oxidative stress, promoting cell proliferation and collagen deposition. This work not only provides a delivery vector for bioactive molecules, but also develops a kind of nanocomposite with the property of promoting wound healing.

Exploring protein-ligand binding affinity prediction with electron density-based geometric deep learning

Rational structure-based drug design relies on accurate predictions of protein-ligand binding affinity from structural molecular information. Although deep learning-based methods for predicting binding affinity have shown promise in computational drug design, certain approaches have faced criticism for their potential to inadequately capture the fundamental physical interactions between ligands and their macromolecular targets or for being susceptible to dataset biases. Herein, we propose to include bond-critical points based on the electron density of a protein-ligand complex as a fundamental physical representation of protein-ligand interactions. Employing a geometric deep learning model, we explore the usefulness of these bond-critical points to predict absolute binding affinities of protein-ligand complexes, benchmark model performance against existing methods, and provide a critical analysis of this new approach. The models achieved root-mean-squared errors of 1.4-1.8 log units on the PDBbind dataset, and 1.0-1.7 log units on the PDE10A dataset, not indicating significant advantages over benchmark methods, and thus rendering the utility of electron density for deep learning models context-dependent. The relationship between intermolecular electron density and corresponding binding affinity was analyzed, and Pearson correlation coefficients r > 0.7 were obtained for several macromolecular targets.

Identification of a hub gene VCL for atherosclerotic plaques and discovery of potential therapeutic targets by molecular docking

BACKGROUND

Atherosclerosis (AS) is a pathology factor for cardiovascular diseases and instability of atherosclerotic plaques contributes to acute coronary events. This study identified a hub gene VCL for atherosclerotic plaques and discovered its potential therapeutic targets for atherosclerotic plaques.

METHODS

Differential expressed genes (DEGs) were screened between unstable and stable plaques from GSE120521 dataset and then used for construction of a protein-protein interactions (PPI) network. Through topological analysis, hub genes were identified within this PPI network, followed by construction of a diagnostic model. GSE41571 dataset was utilized to validate the diagnostic model. A key hub gene was identified and its association with immune characteristics and pathways were further investigated. Molecular docking and molecular dynamics (MD) simulation were employed to discover potential therapeutic targets.

RESULTS

According to the PPI network, 3 tightly connected protein clusters were found. Topological analysis identified the top 5 hub genes, Vinculin (VCL), Dystrophin (DMD), Actin alpha 2 (ACTA2), Filamin A (FLNA), and transgelin (TAGLN). Among these hub genes, VCL had the highest diagnostic value. VCL was selected for further analysis and we found that VCL was negatively correlated with immune score and AS-related inflammatory pathways. Next, we identified 408 genes that were highly correlated with VCL and determined potential drug candidates. The results from molecular docking and MD simulation showed compound DB07117 combined with VCL protein stably, the binding energy is -7.7 kcal/mol, indicating that compound DB07117 was a potential inhibitor of VCL protein.

CONCLUSION

This study identified VCL as a key gene for atherosclerotic plaques and provides a potential therapeutic target of VCL for the treatment of atherosclerotic plaques.

Navigating the Frontiers of Machine Learning in Neurodegenerative Disease Therapeutics

Recent advances in machine learning hold tremendous potential for enhancing the way we develop new medicines. Over the years, machine learning has been adopted in nearly all facets of drug discovery, including patient stratification, lead discovery, biomarker development, and clinical trial design. In this review, we will discuss the latest developments linking machine learning and CNS drug discovery. While machine learning has aided our understanding of chronic diseases like Alzheimer's disease and Parkinson's disease, only modest effective therapies currently exist. We highlight promising new efforts led by academia and emerging biotech companies to leverage machine learning for exploring new therapies. These approaches aim to not only accelerate drug development but to improve the detection and treatment of neurodegenerative diseases.

Larvicidal potential, antimicrobial properties and molecular docking analysis of Egyptian Mint (Mentha rotundifolia) against Culex pipiens L. (Diptera: Culicidae) and Midgut-borne Staphylococcus aureus

Mosquitoes prefer stagnant areas near hospitals to live and easily spread pathogenic bacteria. Our current study aims to isolate multidrug-resistant (MDR) Staphylococcus aureus isolates from midguts of Mosquito Culex pipiens and study the potential of mint as a biocontrol strategy against C. pipiens larvae and their midgut-borne S. aureus. Samples of the third and fourth larval instars of C. pipiens were collected from water ponds around three Cairo hospitals. Ciprofloxacin, gentamycin and tetracycline, as well as various concentrations of mint leaf extract (MLE) were tested for antibiotic susceptibility. Sixty-five isolates were obtained and showed antibiotic resistance to tetracycline, gentamycin, ciprofloxacin, and undiluted MLE with resistant percentages (%) of 27.69, 30.76, 17.46, and 23.08%, respectively. Undiluted MLE inhibited 61.53% of the multidrug S. aureus isolates, whereas it couldn't inhibit any of these isolates at dilutions less than 50 μg/mL. The MIC of MLE was ≤ 700 µg/mL, while the MIC of the antibiotics ranged from 0.25 to 5.0 µg/mL for the three antibiotics. The most inhibited S. aureus isolate was identified by 16SrRNA sequencing approach and registered in GenBank as S. aureus MICBURN with gene accession number OQ766965. MLE killed all larval stages after 72 h of exposure, with mortality (%) reaching 93.33 and 100% causing external hair loss, breakage of the outer cuticle epithelial layer of the abdomen, and larvae shrinkage. Histopathology of treated larvae showed destruction of all midgut cells and organelles. Gas chromatography (GC) of MLE revealed that menthol extract (35.92%) was the largest active ingredient, followed by menthone (19.85%), D-Carvone (15.46%), Pulegone (5.0579%). Docking analysis confirmed that alpha guanine and cadinol had the highest binding affinity to both predicted active sites of Culex pipiens acetylcholinesterase. As a result, alpha-guanine and cadinol might have a role as acetylcholinesterase inhibitors.

Cobdock: an accurate and practical machine learning-based consensus blind docking method

Probing the surface of proteins to predict the binding site and binding affinity for a given small molecule is a critical but challenging task in drug discovery. Blind docking addresses this issue by performing docking on binding regions randomly sampled from the entire protein surface. However, compared with local docking, blind docking is less accurate and reliable because the docking space is too largetly sampled. Cavity detection-guided blind docking methods improved the accuracy by using cavity detection (also known as binding site detection) tools to guide the docking procedure. However, it is worth noting that the performance of these methods heavily relies on the quality of the cavity detection tool. This constraint, namely the dependence on a single cavity detection tool, significantly impacts the overall performance of cavity detection-guided methods. To overcome this limitation, we proposed Consensus Blind Dock (CoBDock), a novel blind, parallel docking method that uses machine learning algorithms to integrate docking and cavity detection results to improve not only binding site identification but also pose prediction accuracy. Our experiments on several datasets, including PDBBind 2020, ADS, MTi, DUD-E, and CASF-2016, showed that CoBDock has better binding site and binding mode performance than other state-of-the-art cavity detector tools and blind docking methods.

Antioxidant, antitumoral, antimetastatic effect and inhibition of collagenase enzyme activity of Eleutherine bulbosa (Dayak onion) extract: In vitro, in vivo and in silico approaches

ETHNOPHARMACOLOGICAL RELEVANCE

Eleutherine bulbosa (Mill.) Urb., known in Brazil as "marupazinho", is a medicinal plant native to the Amazon region. The bulbs of this species are traditionally used in the form of tea or consumed in natura (salads) for the treatment of hypertension, diabetes, gastrointestinal problems, breast cancer, and female fertility. It has been reported that this species possess cytotoxic compounds with anticancer action and limited underlying mechanisms.

AIM OF THE STUDY

This study aimed to analyze extract of E. bulbosa bulbs and evaluate antioxidant activity, antitumor and antimetastatic effects against murine B16F10-Nex2 melanoma cells, and collagenase inhibitory activity by in vitro, in vivo, and in silico approaches.

MATERIALS AND METHODS

Determination of total polyphenols, flavonoids and anthocyanins content were performed. In addition, high performance liquid chromatography-mass spectrometry (HPLC-MS) was carried out to identify phytoconstituents from extract. Antioxidant evaluation was performed using DPPH radical scavenging, ferric ion reducing antioxidant power (FRAP), Thiobarbituric acid reactive species (TBARS) and nitric oxide (NO) tests. Antitumoral and antimetastatic activities of extract on murine B16F10-Nex2 melanoma cells were determined and inhibitory activity on collagenase was evaluated. Molecular interactions between compounds and DNA or collagenase was evaluated by molecular docking analyses.

RESULTS

Phytochemical evaluation demonstrated the presence of polyphenols, flavonoids and anthocyanins, and HPLC-MS identified the major presence of eleutherin, isoeleutherin and eleutherinol. Antioxidant evaluation showed that the extract present significant activity in all methods evaluated. In silico assay demonstrated interaction between bioactive compounds and DNA or collagenase. In addition, extract exhibited antitumor and antimetastatic actions promoted by melanoma cells and showed collagenase inhibitory activity.

CONCLUSIONS

The results showed that E. bulbosa bulb extract contains bioactive compounds as flavonoids, anthocyanins and quinones of which may be responsible for the antioxidant, antitumor, antimetastatic and collagenase enzyme inhibitory activity observed in this study by in vivo, in vitro and in silico bioassays.

The Odorant-Binding Protein 1 Mediates the Foraging Behavior of Grapholita molesta Larvae

Since eggs are laid directly on fruit skin, it is typically believed that food odor has little impact on the foraging of Grapholita molesta larvae. It is crucial to note that larvae that hatch on twigs and leaves could need some sort of identification system when foraging. Here, 22 GmolOBP genes were identified from the G. molesta larval transcriptome via the comparison of conserved domain and homology in the protein level. GmolOBP1 had strong affinities for important pear-fruit volatiles, which caused larvae strong behavioral responses. However, after GmolOBP1 silencing, the larvae lost their attraction to methyl salicylate, α-farnesene, butyl acetate, ethyl butanoate, and ethyl hexanoate, and the effects of larvae seeking various pears were significantly reduced. Consequently, GmolOBP1 was required for the reception of pear volatiles and was involved in mediating how G. molesta larvae foraged. Our research revealed the GmolOBP1 foraging signal recognition mechanism as well as potential molecular targets for field pest management.

Prop3D: A flexible, Python-based platform for machine learning with protein structural properties and biophysical data

BACKGROUND

Machine learning (ML) has a rich history in structural bioinformatics, and modern approaches, such as deep learning, are revolutionizing our knowledge of the subtle relationships between biomolecular sequence, structure, function, dynamics and evolution. As with any advance that rests upon statistical learning approaches, the recent progress in biomolecular sciences is enabled by the availability of vast volumes of sufficiently-variable data. To be useful, such data must be well-structured, machine-readable, intelligible and manipulable. These and related requirements pose challenges that become especially acute at the computational scales typical in ML. Furthermore, in structural bioinformatics such data generally relate to protein three-dimensional (3D) structures, which are inherently more complex than sequence-based data. A significant and recurring challenge concerns the creation of large, high-quality, openly-accessible datasets that can be used for specific training and benchmarking tasks in ML pipelines for predictive modeling projects, along with reproducible splits for training and testing.

RESULTS

Here, we report 'Prop3D', a platform that allows for the creation, sharing and extensible reuse of libraries of protein domains, featurized with biophysical and evolutionary properties that can range from detailed, atomically-resolved physicochemical quantities (e.g., electrostatics) to coarser, residue-level features (e.g., phylogenetic conservation). As a community resource, we also supply a 'Prop3D-20sf' protein dataset, obtained by applying our approach to CATH . We have developed and deployed the Prop3D framework, both in the cloud and on local HPC resources, to systematically and reproducibly create comprehensive datasets via the Highly Scalable Data Service ( HSDS ). Our datasets are freely accessible via a public HSDS instance, or they can be used with accompanying Python wrappers for popular ML frameworks.

CONCLUSION

Prop3D and its associated Prop3D-20sf dataset can be of broad utility in at least three ways. Firstly, the Prop3D workflow code can be customized and deployed on various cloud-based compute platforms, with scalability achieved largely by saving the results to distributed HDF5 files via HSDS . Secondly, the linked Prop3D-20sf dataset provides a hand-crafted, already-featurized dataset of protein domains for 20 highly-populated CATH families; importantly, provision of this pre-computed resource can aid the more efficient development (and reproducible deployment) of ML pipelines. Thirdly, Prop3D-20sf's construction explicitly takes into account (in creating datasets and data-splits) the enigma of 'data leakage', stemming from the evolutionary relationships between proteins.

Recent Deep Learning Applications to Structure-Based Drug Design

Identification and optimization of small molecules that bind to and modulate protein function is a crucial step in the early stages of drug development. For decades, this process has benefitted greatly from the use of computational models that can provide insights into molecular binding affinity and optimization. Over the past several years, various types of deep learning models have shown great potential in improving and enhancing the performance of traditional computational methods. In this chapter, we provide an overview of recent deep learning-based developments with applications in drug discovery. We classify these methods into four subcategories dependent on the task each method is aiming to solve. For each subcategory, we provide the general framework of the approach and discuss individual methods.

LIMD2 is the Signature of Cell Aging-immune/Inflammation in Acute Myocardial Infarction

BACKGROUND

Acute myocardial infarction (AMI) is an age-dependent cardiovascular disease in which cell aging, immunity, and inflammatory factors alter the course; however, cell aging-immune/inflammation signatures in AMI have not been investigated.

METHODS

Based on the GEO database to obtain microRNA (miRNA) sequencing, mRNA sequencing and single-cell sequencing data, and utilizing the Seurat package to identify AMI-associated cellular subpopulations. Subsequently, differentially expressed miRNAs and mRNAs were screened to establish a network of competing endogenous RNAs (ceRNAs). Senescence and immunity scores were calculated by single sample gene set enrichment analysis (ssGSEA), ESTIMATE and CIBERSORT algorithms, and the Hmisc package was used to screen for genes with the highest correlation with senescence and immunity scores. Finally, protein-protein interaction (PPI) and molecular docking analyses were performed to predict potential therapeutic agents for the treatment of AMI.

RESULTS

Four cell types (Macrophage, Fibroblast, Endothelial cells, CD8 T cells) were identified in AMI, and CD8 T cells exhibited the lowest cell aging activity. A ceRNA network of miRNAs- mNRA interactions was established based on the overlapping genes in differentially expressed miRNAs (DEmiRNAs) target genes and differentially expressed mRNAs (DEmRNAs). Twenty-four marker genes of CD8 T cells were observed. LIMD2 was identified as cell aging- immune/inflammation-related hub gene in AMI. This study also identified a potential therapeutic network of DB03276-LIMD2-AMI, which showed excellent and stable binding status between DB03276-LIMD2.

CONCLUSION

This study identified LIMD2 as a cell aging-immune/inflammation-related hub gene. The understanding of the pathogenesis and therapeutic mechanisms of AMI was enriched by the ceRNA network and DB03276-LIMD2-LAMI therapeutic network.

Identification and structural characterization of a novel chondroitin sulfate-specific carbohydrate-binding module: The first member of a new family, CBM100

Chondroitin sulfate is a biologically and commercially important polysaccharide with a variety of applications. Carbohydrate-binding module (CBM) is an important class of carbohydrate-binding protein, which could be utilized as a promising tool for the applications of polysaccharides. In the present study, an unknown function domain was explored from a putative chondroitin sulfate lyase in PL29 family. Recombinant PhCBM100 demonstrated binding capacity to chondroitin sulfates with Ka values of 2.1 ± 0.2 × 106 M-1 and 6.0 ± 0.1 × 106 M-1 to chondroitin sulfate A and chondroitin sulfate C, respectively. The 1.55 Å resolution X-ray crystal structure of PhCBM100 exhibited a β-sandwich fold formed by two antiparallel β-sheets. A binding groove in PhCBM100 interacting with chondroitin sulfate was subsequently identified, and the potential of PhCBM100 for visualization of chondroitin sulfate was evaluated. PhCBM100 is the first characterized chondroitin sulfate-specific CBM. The novelty of PhCBM100 proposed a new CBM family of CBM100.

Genome-Wide Characterization of Berberine Bridge Enzyme Gene Family in Wheat (Triticum aestivum L.) and the Positive Regulatory Role of TaBBE64 in Response to Wheat Stripe Rust

Berberine bridge enzymes (BBEs), functioning as carbonate oxidases, enhance disease resistance in Arabidopsis and tobacco. However, the understanding of BBEs' role in monocots against pathogens remains limited. This study identified 81 TaBBEs with FAD_binding_4 and BBE domains. Phylogenetic analysis revealed a separation of the BBE gene family between monocots and dicots. Notably, RNA-seq showed TaBBE64's significant induction by both pathogen-associated molecular pattern treatment and Puccinia striiformis f. sp. tritici (Pst) infection at early stages. Subcellular localization revealed TaBBE64 at the cytoplasmic membrane. Knocking down TaBBE64 compromised wheat's Pst resistance, reducing reactive oxygen species and promoting fungal growth, confirming its positive role. Molecular docking and enzyme activity assays confirmed TaBBE64's glucose oxidation to produce H2O2. Since Pst relies on living wheat cells for carbohydrate absorption, TaBBE64's promotion of glucose oxidation limits fungal growth and resists pathogen infection. This study sheds light on BBEs' role in wheat resistance against biotrophic fungi.

Identification of novel umami peptides from yeast extract and the mechanism against T1R1/T1R3

Yeast extract was separated by using ultrafiltration, gel filtration chromatography, and preparative high-performance liquid chromatography for analyzing the umami mechanism. 13 kinds of umami peptides were screened out from 73 kinds of peptides which were identified in yeast extract using nanoscale ultra-performance liquid chromatography-tandem mass spectrometry and virtual screening. The umami peptides were found to have a threshold range of 0.07-0.61 mM. DWTDDVEAR exhibited a strong umami taste with a pronounced enhancement effect for monosodium glutamate. Molecular docking studies revealed that specific amino acid residues in the T1R1 subunit, including Arg316, Ser401, and Asp315, played a critical role in the umami perception with these peptides. Overall, the study highlights the potential of natural flavor enhancers and provides insights into the mechanism of umami taste perception.

Multiomics Analysis Reveals the Genetic Basis of Volatile Terpenoid Formation in Oolong Tea

Oolong tea has gained great popularity in China due to its pleasant floral and fruity aromas. Although numerous studies have investigated the aroma differences across various tea cultivars, the genetic mechanism is unclear. This study performed multiomics analysis of three varieties suitable for oolong tea and three others with different processing suitability. Our analysis revealed that oolong tea varieties contained higher levels of cadinane sesquiterpenoids. PanTFBS was developed to identify variants of transcription factor binding sites (TFBSs). We found that the CsDCS gene had two TFBS variants in the promoter sequence and a single nucleotide polymorphism (SNP) in the coding sequence. Integrating data on genetic variations, gene expression, and protein-binding sites indicated that CsDCS might be a pivotal gene involved in the biosynthesis of cadinane sesquiterpenoids. These findings advance our understanding of the genetic factors involved in the aroma formation of oolong tea and offer insights into the enhancement of tea aroma.

The gut-joint axis mediates the TNF-induced RA process and PBMT therapeutic effects through the metabolites of gut microbiota

The gut-joint axis, one of the mechanisms that mediates the onset and progression of joint and related diseases through gut microbiota, and shows the potential as therapeutic target. A variety of drugs exert therapeutic effects on rheumatoid arthritis (RA) through the gut-joint axis. However, the anti-inflammatory and immunomodulatory effect of novel photobiomodulatory therapy (PBMT) on RA need further validation and the involvement of gut-joint axis in this process remains unknown. The present study demonstrated the beneficial effects of PBMT on RA, where we found the restoration of gut microbiota homeostasis, and the related key pathways and metabolites after PBMT. We also discovered that the therapeutic effects of PBMT on RA mainly through the gut-joint axis, in which the amino acid metabolites (Alanine and N-acetyl aspartate) play the key role and rely on the activity of metabolic enzymes in the target organs. Together, the results prove that the metabolites of amino acid from gut microbiota mediate the regulation effect on the gut-joint axis and the therapeutic effect on rheumatoid arthritis of PBMT.

GPCR molecular dynamics forecasting using recurrent neural networks

G protein-coupled receptors (GPCRs) are a large superfamily of cell membrane proteins that play an important physiological role as transmitters of extracellular signals. Signal transmission through the cell membrane depends on conformational changes in the transmembrane region of the receptor, which makes the investigation of the dynamics in these regions particularly relevant. Molecular dynamics (MD) simulations provide a wealth of data about the structure, dynamics, and physiological function of biological macromolecules by modelling the interactions between their atomic constituents. In this study, a Recurrent and Convolutional Neural Network (RNN) model, namely Long Short-Term Memory (LSTM), is used to predict the dynamics of two GPCR states and three specific simulations of each one, through their activation path and focussing on specific receptor regions. Active and inactive states of the GPCRs are analysed in six scenarios involving APO, Full Agonist (BI 167107) and Partial Inverse Agonist (carazolol) of the receptor. Four Machine Learning models with increasing complexity in terms of neural network architecture are evaluated, and their results discussed. The best method achieves an overall RMSD lower than 0.139 Å and the transmembrane helices are the regions showing the minimum prediction errors and minimum relative movements of the protein.

Unraveling viral drug targets: a deep learning-based approach for the identification of potential binding sites

The coronavirus disease 2019 (COVID-19) pandemic has spurred a wide range of approaches to control and combat the disease. However, selecting an effective antiviral drug target remains a time-consuming challenge. Computational methods offer a promising solution by efficiently reducing the number of candidates. In this study, we propose a structure- and deep learning-based approach that identifies vulnerable regions in viral proteins corresponding to drug binding sites. Our approach takes into account the protein dynamics, accessibility and mutability of the binding site and the putative mechanism of action of the drug. We applied this technique to validate drug targeting toward severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike glycoprotein S. Our findings reveal a conformation- and oligomer-specific glycan-free binding site proximal to the receptor binding domain. This site comprises topologically important amino acid residues. Molecular dynamics simulations of Spike in complex with candidate drug molecules bound to the potential binding sites indicate an equilibrium shifted toward the inactive conformation compared with drug-free simulations. Small molecules targeting this binding site have the potential to prevent the closed-to-open conformational transition of Spike, thereby allosterically inhibiting its interaction with human angiotensin-converting enzyme 2 receptor. Using a pseudotyped virus-based assay with a SARS-CoV-2 neutralizing antibody, we identified a set of hit compounds that exhibited inhibition at micromolar concentrations.

Prediction of protein-ligand binding affinity with deep learning

The prediction of binding affinities between target proteins and small molecule drugs is essential for speeding up the drug research and design process. To attain precise and effective affinity prediction, computer-aided methods are employed in the drug discovery pipeline. In the last decade, a variety of computational methods has been developed, with deep learning being the most commonly used approach. We have gathered several deep learning methods and classified them into convolutional neural networks (CNNs), graph neural networks (GNNs), and Transformers for analysis and discussion. Initially, we conducted an analysis of the different deep learning methods, focusing on their feature construction and model architecture. We discussed the advantages and disadvantages of each model. Subsequently, we conducted experiments using four deep learning methods on the PDBbind v.2016 core set. We evaluated their prediction capabilities in various affinity intervals and statistically and visually analyzed the samples of correct and incorrect predictions for each model. Through visual analysis, we attempted to combine the strengths of the four models to improve the Root Mean Square Error (RMSE) of predicted affinities by 1.6% (reducing the absolute value to 1.101) and the Pearson Correlation Coefficient (R) by 2.9% (increasing the absolute value to 0.894) compared to the current state-of-the-art method. Lastly, we discussed the challenges faced by current deep learning methods in affinity prediction and proposed potential solutions to address these issues.

Synthesis and DFT calculations of metal(II) oxime complexes bearing cysteine as coligand and investigation of their biological evolutions in vitro and in silico

New complexes with the formula of [ML(Cys)(H2O)2] were obtained as a result of the reaction between the oxime ligand [HL: 4-(4-bromophenylaminoisonitrosoacetyl)biphenyl], cysteine (Cys), and the metal(II) salts (Mn, Ni, Co, Zn, Cu). The newly synthesized compounds were characterized using conventional techniques such as molar conductance, magnetic measurements, elemental analysis, infrared spectroscopy, and thermal analysis (TGA/DTA). Based on the conductivity measurements in DMF, it was determined that the complexes were non-electrolytes. The TGA/DTA analysis was performed to examine the thermal stability and degradation behavior of all samples, and results demonstrated that metal oxides or sulfides formed as a result of the decompositions. In conjunction with other data obtained, the elemental analysis confirmed the octahedral coordination of the complexes with deprotonated oxime (O, O-donor) and amino acid (N, S-donor) ligands and two coordinated waters. The compounds' optimized geometries, molecular electrostatic potential diagrams, and frontier molecular orbitals were computed at the DFT/B3LYP level using the 6-311 G(d,p) and LANL2DZ basis sets. The antibacterial and DNA cleavage activities of all synthesized compounds were also screened, and molecular docking simulations were performed. According to the results of molecular docking studies conducted with three different proteins, the best interaction was found to be between HL-1HNJ with a binding energy of -9.5 kcal/mol. The stability of the HL-1HNJ complex was also verified by a molecular dynamics simulation performed for 50 ns.Communicated by Ramaswamy H. Sarma.

Dynamic characterization and interpretation for protein-RNA interactions across diverse cellular conditions using HDRNet

RNA-binding proteins play crucial roles in the regulation of gene expression, and understanding the interactions between RNAs and RBPs in distinct cellular conditions forms the basis for comprehending the underlying RNA function. However, current computational methods pose challenges to the cross-prediction of RNA-protein binding events across diverse cell lines and tissue contexts. Here, we develop HDRNet, an end-to-end deep learning-based framework to precisely predict dynamic RBP binding events under diverse cellular conditions. Our results demonstrate that HDRNet can accurately and efficiently identify binding sites, particularly for dynamic prediction, outperforming other state-of-the-art models on 261 linear RNA datasets from both eCLIP and CLIP-seq, supplemented with additional tissue data. Moreover, we conduct motif and interpretation analyses to provide fresh insights into the pathological mechanisms underlying RNA-RBP interactions from various perspectives. Our functional genomic analysis further explores the gene-human disease associations, uncovering previously uncharacterized observations for a broad range of genetic disorders.

Structure of the human heparan-α-glucosaminide N-acetyltransferase (HGSNAT)

Degradation of heparan sulfate (HS), a glycosaminoglycan (GAG) comprised of repeating units of N-acetylglucosamine and glucuronic acid, begins in the cytosol and is completed in the lysosomes. Acetylation of the terminal non-reducing amino group of α-D-glucosamine of HS is essential for its complete breakdown into monosaccharides and free sulfate. Heparan-α-glucosaminide N-acetyltransferase (HGSNAT), a resident of the lysosomal membrane, catalyzes this essential acetylation reaction by accepting and transferring the acetyl group from cytosolic acetyl-CoA to terminal α-D-glucosamine of HS in the lysosomal lumen. Mutation-induced dysfunction in HGSNAT causes abnormal accumulation of HS within the lysosomes and leads to an autosomal recessive neurodegenerative lysosomal storage disorder called mucopolysaccharidosis IIIC (MPS IIIC). There are no approved drugs or treatment strategies to cure or manage the symptoms of, MPS IIIC. Here, we use cryo-electron microscopy (cryo-EM) to determine a high-resolution structure of the HGSNAT-acetyl-CoA complex in an open-to-lumen conformation, the first step in HGSNAT catalyzed acetyltransferase reaction. In addition, we map the known MPS IIIC mutations onto the structure and elucidate the molecular basis for mutation-induced HGSNAT dysfunction.