Definition of germ layer cell lineage alternative splicing programs reveals a critical role for Quaking in specifying cardiac cell fate

Alternative splicing is critical for development; however, its role in the specification of the three embryonic germ layers is poorly understood. By performing RNA-Seq on human embryonic stem cells (hESCs) and derived definitive endoderm, cardiac mesoderm, and ectoderm cell lineages, we detect distinct alternative splicing programs associated with each lineage. The most prominent splicing program differences are observed between definitive endoderm and cardiac mesoderm. Integrative multi-omics analyses link each program with lineage-enriched RNA binding protein regulators, and further suggest a widespread role for Quaking (QKI) in the specification of cardiac mesoderm. Remarkably, knockout of QKI disrupts the cardiac mesoderm-associated alternative splicing program and formation of myocytes. These changes arise in part through reduced expression of BIN1 splice variants linked to cardiac development. Mechanistically, we find that QKI represses inclusion of exon 7 in BIN1 pre-mRNA via an exonic ACUAA motif, and this is concomitant with intron removal and cleavage from chromatin. Collectively, our results uncover alternative splicing programs associated with the three germ lineages and demonstrate an important role for QKI in the formation of cardiac mesoderm.     

Chromosome-level genome assembly and characterization of Sophora Japonica

Sophora japonica is a medium-size deciduous tree belonging to Leguminosae family and famous for its high ecological, economic and medicinal value. Here, we reveal a draft genome of S. japonica, which was ∼511.49 Mb long (contig N50 size of 17.34 Mb) based on Illumina, Nanopore and Hi-C data. We reliably assembled 110 contigs into 14 chromosomes, representing 91.62% of the total genome, with an improved N50 size of 31.32 Mb based on Hi-C data. Further investigation identified 271.76 Mb (53.13%) of repetitive sequences and 31,000 protein-coding genes, of which 30,721 (99.1%) were functionally annotated. Phylogenetic analysis indicates that S. japonica separated from Arabidopsis thaliana and Glycine max ∼107.53 and 61.24 million years ago, respectively. We detected evidence of species-specific and common-legume whole-genome duplication events in S. japonica. We further found that multiple TF families (e.g. BBX and PAL) have expanded in S. japonica, which might have led to its enhanced tolerance to abiotic stress. In addition, S. japonica harbours more genes involved in the lignin and cellulose biosynthesis pathways than the other two species. Finally, population genomic analyses revealed no obvious differentiation among geographical groups and the effective population size continuously declined since 2 Ma. Our genomic data provide a powerful comparative framework to study the adaptation, evolution and active ingredients biosynthesis in S. japonica. More importantly, our high-quality S. japonica genome is important for elucidating the biosynthesis of its main bioactive components, and improving its production and/or processing.     

DeepMHADTA: Prediction of Drug-Target Binding Affinity Using Multi-Head Self-Attention and Convolutional Neural Network

Drug-target interactions provide insight into the drug-side effects and drug repositioning. However, wet-lab biochemical experiments are time-consuming and labor-intensive, and are insufficient to meet the pressing demand for drug research and development. With the rapid advancement of deep learning, computational methods are increasingly applied to screen drug-target interactions. Many methods consider this problem as a binary classification task (binding or not), but ignore the quantitative binding affinity. In this paper, we propose a new end-to-end deep learning method called DeepMHADTA, which uses the multi-head self-attention mechanism in a deep residual network to predict drug-target binding affinity. On two benchmark datasets, our method outperformed several current state-of-the-art methods in terms of multiple performance measures, including mean square error (MSE), consistency index (CI), rm2, and PR curve area (AUPR). The results demonstrated that our method achieved better performance in predicting the drug–target binding affinity.     

SARS-CoV-2 envelope protein causes acute respiratory distress syndrome (ARDS)-like pathological damages and constitutes an antiviral target

Cytokine storm and multi-organ failure are the main causes of SARS-CoV-2-related death. However, the origin of excessive damages caused by SARS-CoV-2 remains largely unknown. Here we show that the SARS-CoV-2 envelope (2-E) protein alone is able to cause acute respiratory distress syndrome (ARDS)-like damages in vitro and in vivo. 2-E proteins were found to form a type of pH-sensitive cation channels in bilayer lipid membranes. As observed in SARS-CoV-2-infected cells, heterologous expression of 2-E channels induced rapid cell death in various susceptible cell types and robust secretion of cytokines and chemokines in macrophages. Intravenous administration of purified 2-E protein into mice caused ARDS-like pathological damages in lung and spleen. A dominant negative mutation lowering 2-E channel activity attenuated cell death and SARS-CoV-2 production. Newly identified channel inhibitors exhibited potent anti-SARS-CoV-2 activity and excellent cell protective activity in vitro and these activities were positively correlated with inhibition of 2-E channel. Importantly, prophylactic and therapeutic administration of the channel inhibitor effectively reduced both the viral load and secretion of inflammation cytokines in lungs of SARS-CoV-2-infected transgenic mice expressing human angiotensin-converting enzyme 2 (hACE-2). Our study supports that 2-E is a promising drug target against SARS-CoV-2.Subject terms: Cell death, Molecular biology     

合成己酸乙酯脂肪酶产生菌的筛选及产酶条件

从２７株脂肪酶产生菌中筛选到能由乙醇和己酸合成己酸乙酯的菌株８株。其中Ｒｈｉｚｏｐｕｓｓｐ．Ｈ－３菌株脂肪酶活力为５０－６０ｕ／ｍｌ,全细胞在有机溶剂中的酯化率可达己酸的９１％。Ｈ３产酶的最适碳源为淀粉或葡萄糖。６％黄豆饼粉加４％蛋白陈复合氮源有利于酶活力的增加。