Genome-Wide Screening for Dominant Modifiers in Drosophila Identified New Cluster of Genes Involved in BCR-ABL Signalling and CML Progression

Despite the role of Bcr-Abl in the pathogenesis of Chronic Myeloid Leukaemia (CML) is well established, the mechanisms responsible for CML progression remain largely unknown. The aim of the study was to perform a genome-wide screening to identify new genes and pathways leading to CML progression. We performed a genome-wide genetic screening using our set-up model of human p210 Bcr-Abl transgenic Drosophila melanogaster (Dm) in which the expression of hBcr-Abl in a tissue specific manner is able to induce a severe eye-glazed phenotype or the formation of melanotic tumors, (clusters of hemocytes) when expressed into the fly lymph gland which represent the Dm hematopoietic system. A wide modifier screening of the whole fly genome containing approx 14.000 genes was performed using 278 fly stocks commercially available and carrying well characterized chromosome deletions. The resulting progeny was screened using the eye phenotype as first read-out system. Furthermore each deletion responsible for phenotype changes was analyzed either by expressing it into lymph gland as second read-out system, in order to analyze their function into a haematopoietic background and to exclude genes involved in eye development, such as genes able to modify the eye phenotype even without being directly involved in Bcr-Abl oncogenic signalling (false positives). Data obtained from primary and secondary screen were first analyzed using the Gene Ontology software. These results were compared with gene expression signatures of CML from Microarray data. As final point, the identified candidate genes were tested and validated analyzing either BM or PB samples from CML patients and healthy donors. 14.000 Dm genes were analyzed for their capability to genetically interact with hBcr-Abl in the fly model. The analysis of eye/lymph gland-phenotypes in the progeny obtained from screening crosses, shows a first group of flies (38%) displaying a more aggressive phenotype since they lack genes encoding for hBcr-Abl negative regulators and a second group (32%) showing a mild phenotype due to the absence of genes involved in the oncogenic signalling. We found that 42% of the 4000 Dm genes mapping in the deleted regions able to modify Bcr-Abl phenotype, displayed a known human counterpart. GeneOntology profiles of these genes included oncogenes, tumor suppression genes and human genes encoding proteins involved in the regulation of transcriptions, signal transduction, proliferation and cell growth, differentiation, apoptosis and splicing processes. Moreover, a computational comparison of our results with gene expression signatures of CML from Microarray data, showed only a partially overlap between genes identified in fly screen and genes obtained from Microarray analysis. The 72% of identified genes in fact was not known to be involved in human leukaemia. However, further confirmation of our findings into fly comes from the validation in human samples in which 1250 genes were found to be significantly associated with human CML; among these genes, we found not only an alteration of their expression profiles in CML patients with respect to the healthy donors, but also protein alterations, such as expression of different splicing forms or misplaced proteins, suggesting that Dm screening is a valid approach able to identify not only differentially expressed genes but also specific pathways and genes otherwise altered by hBcr-Abl. In conclusion, the identification of these genes allows identifying of the changes occurring in CML at the genomic level and gives deeper insights into the molecular basis of the disease; moreover this study point to specific gene pathways that might represent new targets for therapy in CML in order to prevent or overcome resistance and progression