ToxMPNN: A deep learning model for small molecule toxicity prediction.

Machine learning (ML) has shown a great promise in predicting toxicity of small molecules. However, the availability of data for such predictions is often limited. Because of the unsatisfactory performance of models trained on a single toxicity endpoint, we collected toxic small molecules with multiple toxicity endpoints from previous study. The dataset comprises 27 toxic endpoints categorized into seven toxicity classes, namely, carcinogenicity and mutagenicity, acute oral toxicity, respiratory toxicity, irritation and corrosion, cardiotoxicity, CYP450, and endocrine disruption. In addition, a binary classification Common-Toxicity task was added based on the aforementioned dataset. To improve the performance of the models, we added marketed drugs as negative samples. This study presents a toxicity predictive model, ToxMPNN, based on the message passing neural network (MPNN) architecture, aiming to predict the toxicity of small molecules. The results demonstrate that ToxMPNN outperforms other models in capturing toxic features within the molecular structure, resulting in more precise predictions with the ROC_AUC testing score of 0.886 for the Toxicity_drug dataset. Furthermore, it was observed that adding marketed drugs as negative samples not only improves the predictive performance of the binary classification Common-Toxicity task but also enhances the stability of the model prediction. It shows that the graph-based deep learning (DL) algorithms in this study can be used as a trustworthy and effective tool to assess small molecule toxicity in the development of new drugs.

Cardiotoxicity of (-)-borneol, (+)-borneol, and isoborneol in zebrafish embryos is associated with Na+ /K+ -ATPase and Ca2+ -ATPase inhibition.

Borneol is an example of traditional Chinese medicine widely used in Asia. There are different isomers of chiral borneol in the market, but its toxicity and effects need further study. In this study, we used zebrafish embryos to examine the effects of exposure to three isomers of borneol [(-)-borneol, (+)-borneol, and isoborneol] on heart development and the association with Na+ /K+ -ATPase from 4 h post-fertilization (4 hpf). The results showed that the three isomers of borneol increased mortality and decreased hatching rate when the zebrafish embryo developed to 72 hpf. All three isomers of borneol (0.01-1.0 mM) significantly reduced heart rate from 48 to 120 hpf and reduced the expression of genes related to Ca2+ -ATPase (cacna1ab and cacna1da) and Na+ /K+ -ATPase (atp1b2b, atp1a3b, and atp1a2). At the same time, the three isomers of borneol significantly reduced the activities of Ca2+ -ATPase and Na+ /K+ -ATPase at 0.1 to 1.0 mM. (+)-Borneol caused the most significant reduction (p < 0.05), followed by isoborneol and (-)-borneol. Na+ /K+ -ATPase was mainly expressed in otic vesicles and protonephridium. All three isomers of borneol reduced Na+ /K+ -ATPase mRNA expression, but isoborneol was the most significant (p < 0.01). Our results indicated that (+)-borneol was the least toxic of the three isomers while the isoborneol showed the most substantial toxic effect, closely related to effects on Na+ /K+ -ATPase.

Identification of Potential Molecular Targets and Active Ingredients of Mingmu Dihuang Pill for the Treatment of Diabetic Retinopathy Based on Network Pharmacology.

Objective: Mingmu Dihuang Pill (MMDHP) is a traditional Chinese formula that has shown remarkable improvements of dry eyes, tearing, and blurry vision; however, the mechanisms underlying MMDHP treatment for diabetic retinopathy have not been fully understood. This study is aimed at identifying the molecular targets and active ingredients of MMDHP for the treatment of diabetic retinopathy based on network pharmacology. Methods: All active ingredients of MMDHP were retrieved from TCMSP and BATMAN-TCM databases, and the targets of active ingredients of MMDHP were predicted on the SwissTargetPrediction website. Diabetic retinopathy-related target sets were retrieved from GeneCards and OMIM databases, and the intersecting targets between targets of active ingredients of MMDHP and potential therapeutic targets of diabetic retinopathy were collected to generate the traditional Chinese medicine-ingredient-target-diabetic retinopathy network and to create the protein-protein interaction network. In addition, GO terms and KEGG pathway enrichment analyses were performed to identify the potential pathways, and molecular docking was employed to verify the binding of active ingredients of MMDHP to key targets of diabetic retinopathy. Results: Network pharmacology predicted 183 active ingredients and 904 targets from MMDHP, and 203 targets were intersected with the therapeutic targets of diabetic retinopathy. The top 10 hub targets included PIK3RA, TP53, SRC, JUN, HRAS, AKT1, VEGFA, EGFR, ESR1, and PI3KCA. GO terms and KEGG pathway enrichment analyses identified AGE-RAGE, PI3K-AKT, and Rap1 signaling pathways as major pathways involved in MMDHP treatment for diabetic retinopathy. Molecular docking confirmed a good binding affinity of active ingredients of MMDHP, including luteolin, acacetin, naringenin, and alisol B, with AKT1, SRC, and VEGFA as the three key targets of diabetic retinopathy. Conclusion: MMDHP may be effective for the treatment of diabetic retinopathy through active ingredients luteolin, acacetin, naringenin, and alisol B via AKT1, SRC, and VEGFA in AGE-RAGE, PI3K-AKT, and Rap1 signaling pathways.

Effects of Vitexin, a Natural Flavonoid Glycoside, on the Proliferation, Invasion, and Apoptosis of Human U251 Glioblastoma Cells.

Glioblastoma is a highly aggressive brain tumor characterized by high recurrence and poor prognosis. Vitexin has shown activities against esophageal, liver, lung, colorectal, and ovarian cancers; however, there is little knowledge on the activity of vitexin against glioblastoma. This study was therefore designed with aims to examine the effects of vitexin on proliferation, invasion, and apoptosis of human U251 glioblastoma cells and explore the underlying molecular mechanisms using mRNA sequencing and molecular docking. Vitexin was found to inhibit cell proliferation, colony formation, and invasion and promote apoptosis in U251 cells. mRNA sequencing identified 499 differentially expressed genes in vitexin-treated U251 cells relative to controls, including 154 upregulated genes and 345 downregulated genes. Gene ontology (GO) term enrichment analysis revealed that the upregulated genes were most significantly enriched in intrinsic apoptotic signaling pathway and the downregulated genes were most significantly enriched in positive regulation of cell development and positive regulation of locomotion relating to biological processes, endoplasmic reticulum lumen and side of membrane relating to cellular components, and receptor ligand activity and receptor regulator activity relating to molecular functions. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that the upregulated genes were involved in the pathways of transcriptional misregulation in cancer and the downregulated genes were involved in FoxO and JAK/STAT signaling pathways. Western blotting assay revealed that vitexin treatment resulted in reduced p-JAK1, p-JAK3, and p-STAT3 protein expression in U251 cells relative to untreated controls, and molecular docking predicted that vitexin had docking scores of -8.8, -10.8, and -10.5 kJ/mol with STAT3, JAK1, and JAK2, respectively. The results of the present study demonstrate that vitexin inhibits the proliferation and invasion and induces the apoptosis of glioblastoma U251 cells through suppressing the JAK/STAT3 signaling pathway, and vitexin may be a promising potential agent for the chemotherapy of glioblastoma.