Catalytically distinct metabolic enzyme isocitrate dehydrogenase 1 mutants tune phenotype severity in tumor models

Mutations in isocitrate dehydrogenase 1 (IDH1) impart a neomorphic reaction that produces D-2-hydroxyglutarate (D2HG), which can inhibit DNA demethylases to drive tumorigenesis. Mutations affect residue R132 and display distinct catalytic profiles for D2HG production. We show that catalytic efficiency of D2HG production is greater in IDH1 R132Q than R132H mutants, and expression of IDH1 R132Q in cellular and xenograft models leads to higher D2HG concentrations in cells, tumors, and sera compared to R132H. Though expression of IDH1 R132Q leads to hypermethylation in DNA damage pathways, DNA hypomethylation is more notable when compared to IDH1 R132H expression. Transcriptome analysis shows increased expression of many pro-tumor pathways upon expression of IDH1 R132Q versus R132H, including transcripts of EGFR and PI3K signaling pathways. Thus, IDH1 mutants appear to modulate D2HG levels via altered catalysis and are associated with distinct epigenetic and transcriptomic consequences, with higher D2HG levels appearing to be associated with more aggressive tumors. Overall design: To investigate the consequences of gene expression as measured by RNAseq analysis in mouse xenograft tumors generated from human cells stably overexpressing IDH! Wild type (WT), R132H, or R132Q in U87MG cells or in HT1080 cells modified to have the IDH1 R132C allele removed (WT/null alleles, or +/- here). Comparative gene expression profiling analysis of RNAseq data for the resulting tumors are provided. WT, R132H, and R132Q forms of IDH1 contained a C-terminal HA tag and were stably introduced into cells using an LEIH vector that contains an EF1a promoter for high expression in lentiviral based transfection. Female athymic nude mice 6-8 weeks old were purchased from Jackson Labs were used. 27 mince per cell line (U87MG or HT1080) were used, with 9 mice per IDH type (WT, R132H, R132Q). The tumor from individual mice, if formed, were harvested if formed and a randomized subset were processed for RNAseq analysis.