Skip to main content
. 2018 Jun 11;9:2268. doi: 10.1038/s41467-018-04730-7

Fig. 1.

Fig. 1

Arsenic stress elicits global 3′UTR shortening. a Schematic of experimental design. AS, 1 h treatment with 250 µM sodium arsenite; RC, recovery after AS. b Top, western blot analysis of phosphorylated (upper) and total (lower) eIF-2α protein. NT non-treated. Bottom, normalized ratio of amount of phosphorylated eIF-2α to total eIF-2α. Error bars are standard deviation based on two replicates. c Immunocytochemistry analysis of stress granules (SGs) using anti-PABPC1. Scale bar: 10 µm. d Top, schematic of two APA isoforms using proximal PAS (pPAS) or distal PAS (dPAS) in the 3′UTR. The region between the two PASs is named alternative 3′UTR (aUTR). Bottom, scatter plot showing expression change of pPAS isoform (x-axis) and that of dPAS isoform (y-axis) in total RNA after 1 h of AS. Genes with significantly shortened or lengthened 3′UTRs (P < 0.05, Fisher’s exact test) based on two biological replicates are highlighted in blue and red, respectively. e Median relative expression difference (RED) values of treatment samples compared to the NT sample, which reflect the degree of global 3′UTR length changes. Significance of difference (Wilcoxon test) is indicated, with three asterisks indicating P < 0.001, and n.s. P > 0.05. Error bars are standard deviations based on random sampling of data for 20 times. f Relationship between RED and aUTR size. Genes were divided into five bins based on aUTR size. Median RED for each gene bin is shown for AS (red line) or RC (blue line) vs. NT cells. Error bars are standard error of mean. P-values (Wilcoxon test) are based on difference between gene bins 1 and 5. g 3′READS data of an example gene Nmt1, visualized in UCSC genome browser. Expression levels of isoforms are indicated by reads per million (RPM) PAS reads. Note the RefSeq track at the top does not cover the dPAS due to poor annotation. h RT-qPCR analysis of the relative amounts of APA isoforms (x-axis). Two primer sets were used for each gene, targeting a common region to both isoforms and the aUTR, respectively. The log2(aUTR/common) value (y-axis) indicates relative abundance of long vs. short 3′UTR isoforms. Error bars are standard deviation based on two replicates