Gene expression analyses of E14.5 fetal liver Lin-Sca1+ckit+ (LSK) cells from wild-type, β-Catenin-KO, N1IC+ and N1IC+β-Catenin-KO embryos.. Mus musculus

Notch activation is instrumental in the development of most T-cell acute lymphoblastic leukemia (T-ALL) cases, yet Notch mutations alone are not sufficient to recapitulate the full human disease in animal models. Using multiple in vivo and in vitro T-ALL models we here demonstrate that β-Catenin is essential for Notch-driven T-cell leukemic initiation. Transcriptome analyses of leukemic initiating cells revealed a switch in β-Catenin activity that was Notch-context dependent. Moreover, ChIP-seq coupled with RNA-Seq in human Notch-active T-ALL showed that leukemic β-Catenin was independent of canonical LEF/TCF partners, and instead depended on direct association with Notch or ZBTB33/Kaiso for gene activation. The functional relevance of this mechanism is exemplified by the MYC 3´enhancer that requires β-Catenin and Notch1 recruitment to induce MYC expression. Finally, we demonstrate that pharmacological inhibition of β-Catenin with PKF115-584 prevented and partially reverted leukemogenesis induced by active Notch1. These microarray data show the transcriptional activities of N1IC and β-Catenin in wild-type or leukemic initiating cell (LIC) contexts in E14.5 FL LSK cells. Overall design: Study is based on biological triplicates. Total RNA from 5-50k sorted FL LSK cells was isolated with RNEasy Mini or Micro kits (QIAgen). RNA quality was assessed with the Bioanalyzer 2100 PicoChip (Agilent Technologies, Palo Alto, CA) and only RNA samples with RIN >6 was used for microarrays. RNAs were amplified using the Ovation™ Pico WTA System (NuGEN Technologies, San Carlos, CA) and sense transcript cDNA (ST-cDNA) was generated using the WT-Ovation™ Exon Module (NuGEN Technologies). After, ST-cDNA was fragmented and labelled with the FL-Ovation™ cDNA Biotin Module V2 (NuGEN Technologies), and the biotinylated cDNA were hybridized to Affymetrix GeneChip® Mouse Gene 2.0 ST arrays as biological triplicates. Following hybridization, the arrays were washed, stained and scanned to generate CEL files for each array. CEL files were analyzed in R (v3.1.2) with the oligo (v1.30.0) and limma (v3.22.4) packages from Bioconductor, data were normalized using RMA (Robust Multi-array Average) and expression values log2-transformed.