BCR/ABL1-Dependent Transcriptional Response Reveals Enrichment for Genes Involved in Negative Feedback Regulation. Homo sapiens

Philadelphia (Ph) chromosome-positive leukemia is characterized by the BCR/ABL1 fusion protein that affects a wide range of signal transduction pathways. The knowledge about its downstream target genes is, however, still quite limited. To identify novel BCR/ABL1-regulated genes we used global gene expression profiling of several Ph-positive and Ph-negative cell lines treated with imatinib. Following imatinib treatment, the Ph-positive cells showed decreased growth, viability, and reduced phosphorylation of BCR/ABL1 and STAT5. In total, 142 genes were identified as being dependent on BCR/ABL1-mediated signaling, mainly including genes involved in signal transduction, e.g. the JAK/STAT, MAPK, TGFB and insulin signaling pathways, and in regulation of metabolism. Interestingly, BCR/ABL1 was found to activate several genes involved in negative feedback regulation (CISH, SOCS2, SOCS3, PIM1, DUSP6, and TNFAIP3), which may act to indirectly suppress the tumor promoting effects exerted by BCR/ABL1. In addition, several genes identified as deregulated upon BCR/ABL1 expression could be assigned to the TGFB and NFkB signaling pathways, as well as to reflect the metabolic adjustments needed for rapidly growing cells. Apart from providing important pathogenetic insights into BCR/ABL1-mediated leukemogenesis, the present study also provides a number of pathways/individual genes that may provide attractive targets for future development of targeted therapies. Keywords: global gene expression profiling, Ph chromosome, BCR/ABL1, imatinib mesylate Overall design: Five Philadelphia-positive (Ph+) and five Philadelphia-negative (Ph-) cell lines were cultured in the absence or presence of 1 microM imatinib mesylate for 3 and 12 hours. Total RNA was extracted from untreated and treated cell cultures at both time points, resulting in 4 samples per cell line. RNA extraction, labeling, hybridization, washing, scanning, and feature analysis were performed as described (Håkansson et al., 2008, Genes Chromosomes Cancer). Dye-swap labeling was performed in 32 out of 40 samples. In total, 72 slides were hybridized and scanned. The transcriptional response was studied using the mean value of the 3 and 12 hour microarray measurements.