Fig. 2: HuR regulates GAC and KGA isoform levels.
MIA PaCa-2 expression levels of ELAVL1 and GLS from RNA-Seq data (a, mean and S.E.M.) and heatmap (b) with bidirectional clustering of singscores111 from KEGG Metabolic Pathways and scaled ELAVL1 expression. The letter ‘Q’ marks pathways that involve glutamine from (Control n = 4; Knockout n = 3). The dendrograms represent full bidirectional complete clustering based on correlation values. c Chromosome 2 indicating GLS loci and exon structure for KGA and GAC isoforms. Bellow, exon differential expression after ELAVL1 knockdown in MIA PaCa-2 or HeLa cells. Asterisks denote FDR < 0.05. d, above, Western blot of doxycycline-inducible ELAVL1-silenced breast cancer cell lines (MDA-MB-231 and BT549) using two shRNAs, displaying a marked decrease in KGA protein level, accompanied by an increase in GAC. d, below, Western blot of constitutively silenced ELAVL1 breast cancer cell lines, grouped by telaglenastat resistance, which was defined elsewhere46. Arrows indicate KGA and GAC isoforms and HuR-specific bands and western blots repeated at least two times with reproducible results. e Scheme for CRISPR-mediated knock-in of the mKO2-P2A-BleoR cassette into GLS exon 19 to produce the KGA-mKO2 fusion protein. f Increase and decrease in cellular fluorescence after ELAVL1 ectopic expression (bottom) or shELAVL1-mediated transient silencing (top), respectively, in HEK293T knock-in cells, number independent cells evaluated per condition >1433. Representative images on the right. The scale bar is 50 μm. g Cell cycle analysis of BT549 knock-in cells KGA levels (evaluated by mKO2 fluorescence) are enhanced in S-G2/M phases, as expected50,51. Representative images in bottom part, with HiLo lookup table for DNA staining, one cell with higher DNA content and higher KGA content and the opposite example below, confirming the functionality of the system. Box plots represent the interquartile range; the vertical curve is the kernel density of the distribution, and the dark horizontal line denotes the median. Each dot represents an individual cell. Statistical significance derived from bootstrapped105 DEXSeq107 (c) or Two-sided Welch’s t-test (f). *p < 0.05, **p < 0.01, ***p < 0.0001.
