Meta-analysis of whole-genome gene expression datasets assessing the effects of IDH1 and IDH2 mutations in isogenic disease models

Abstract

Mutations in isocitrate dehydrogenase 1 (IDH1) and IDH2 are oncogenic drivers to a variable extent in several tumors, including gliomas, acute myeloid leukemia (AML), cholangiocarcinoma, melanoma, and thyroid carcinoma. The pathobiological effects of these mutations vary considerably, impeding the identification of common expression profiles. We performed an expression meta-analysis between IDH-mutant (IDHmut) and IDH-wild-type (IDHwt) conditions in six human and mouse isogenic disease models. The datasets included colon cancer cells, glioma cells, heart tissue, hepatoblasts, and neural stem cells. Among differentially expressed genes (DEGs), serine protease 23 (PRSS23) was upregulated in four datasets, i.e., in human colon carcinoma cells, mouse heart tissue, mouse neural stem cells, and human glioma cells. Carbonic anhydrase 2 (CA2) and prolyl 3-hydroxylase 2 (P3H2) were upregulated in three datasets, and SOX2 overlapping transcript (SOX2-OT) was downregulated in three datasets. The most significantly overrepresented protein class was termed intercellular signal molecules. An additional DEG set contained genes that were both up- and downregulated in different datasets and included oxidases and extracellular matrix structural proteins as the most significantly overrepresented protein classes. In conclusion, this meta-analysis provides a comprehensive overview of the expression effects of IDH mutations shared between different isogenic disease models. The generated dataset includes biomarkers, e.g., PRSS23 that may gain relevance for further research or clinical applications in IDHmut tumors.

PolyAMiner-Bulk: A Machine Learning Based Bioinformatics Algorithm to Infer and Decode Alternative Polyadenylation Dynamics from bulk RNA-seq data

“…For example, TAMM41, involved in mitochondrial translocator assembly and maintenance, is a representative differential APA gene with a positive PolyAIndex metric, suggesting that this gene is undergoing 3’UTR lengthening in the RBM17 knockdown condition compared to the control condition (Figure 5C) (38, 39). On the other hand, P3H2, which is involved in collagen chain assembly and stability, is a representative differential APA gene with a negative PolyAIndex metric, suggesting that this gene is undergoing 3’UTR shortening in the RBM17 knockdown condition compared to the control condition (Figure 5D) (40, 41). We visualized both genes’ read density – as bulk RNA-seq and pseudo-3’UTR-seq read coverage – and plotted their corresponding density proportions as a heatmap.…”

Moderate maternal nutrient reduction in pregnancy alters fatty acid oxidation and RNA splicing in the nonhuman primate fetal liver

“…[27][28][29][30][31][32] In fact, two of the key genes identified in the networks here -IDH1 and SRSF2are recurrently mutated and are actually definitive of certain cancers. [27][28][29]33,31 The other key gene identified, PRKAR1A, when mutated, leads to overactive protein kinase A which has been found to phosphorylate splicing factors (among many other roles). [34][35][36] Interestingly, PRKAR1A mutations result in Carney complex which is characterized by an increased risk of several types of tumors and significant phenotypic heterogeneity.…”