Fucosyltransferase 8 (FUT8) and core fucose expression in oxidative stress response.

GlycoMaple is a new tool to predict glycan structures based on the expression levels of 950 genes encoding glycan biosynthesis-related enzymes and proteins using RNA-seq data. The antioxidant response, protecting cells from oxidative stress, has been focused on because its activation may relieve pathological conditions, such as neurodegenerative diseases. Genes involved in the antioxidant response are defined within the GO:0006979 category, including 441 human genes. Fifteen genes overlap between the glycan biosynthesis-related genes defined by GlycoMaple and the antioxidant response genes defined by GO:0006979, one of which is FUT8. 5-Hydroxy-4-phenyl-butenolide (5H4PB) extracted from Chinese aromatic vinegar induces the expression of a series of antioxidant response genes that protect cells from oxidative stress via activation of the nuclear factor erythroid 2-related factor 2-antioxidant response element pathway. Here, we show that FUT8 is upregulated in both our RNA-seq data set of 5H4PB-treated cells and publicly available RNA-seq data set of cells treated with another antioxidant, sulforaphane. Applying our RNA-seq data set to GlycoMaple led to a prediction of an increase in the core fucose of N-glycan that was confirmed by flow cytometry using a fucose-binding lectin. These results suggest that FUT8 and core fucose expression may increase upon the antioxidant response.

A Primer on Deep Learning-Based Cellular Image Classification of Changes in the Spatial Distribution of the Golgi Apparatus After Experimental Manipulation.

The visual classification of cell images according to differences in the spatial patterns of subcellular structure is an important methodology in cell and developmental biology. Experimental perturbation of cell function can induce changes in the spatial distribution of organelles and their associated markers or labels. Here, we demonstrate how to achieve accurate, unbiased, high-throughput image classification using an artificial intelligence (AI) algorithm. We show that a convolutional neural network (CNN) algorithm can classify distinct patterns of Golgi images after drug or siRNA treatments, and we review our methods from cell preparation to image acquisition and CNN analysis.