Genome-Wide Analysis of MDS/MPD Disclosed Frequent Homozygous C-Cbl mutations Tightly Associated with 11q-UPD

Myelodysplastic syndromes (MDS) are clonal disorders of hematopoietic progenitors characterized by ineffective hematopoiesis and high propensity to leukemias. Although a number of gene targets have been identified, in many MDS cases, particular genetic targets are unknown. In this study, we performed genome-wide profiling of copy number (CN) abnormalities and allelic imbalances in MDS genomes in order to clarify the distribution of LOH (loss of heterozygosity) and to identify their gene targets. We analyzed a total of 171MDS and MDS/MPD specimens, including 7 RA/RARS, 23 RCMD/RCMD-RS, 6 5q-syndrome, 30 RAEB-1, 40 RAEB-2, 4 therapy related-MDS/AML, 5 MDSu, 17 CMML-1, 16 CMML-2, 24 overt AML, using high-density SNP arrays. The data were analyzed by CNAG/AsCNAR software, which enabled allele-specific CN analysis and sensitive LOH detection. MDS showed characteristic CN profiles in SNP array analysis. Of particular interest is the finding of high frequency of CN-neutral LOH (Uniparental disomy,UPD) observed in 51 of 171 (30%) MDS cases. They preferentially involved 1p, 1q, 4q, 7q, 11q, 17p and other chromosomal segments, which were associated with homozygous mutations of both loss-of-function mutations and gain-of function mutations of tumor suppressor genes and cellular oncogenes, including TP53 (17p UPD), AML1/RUNX1 (21q UPD), Nras and cMPL (1p UPD), JAK-2 (9p UPD), and FLT3 (13q UPD). Next we tried to identify a new gene target in 11q UPD, which was most common UPD region in this study and many of these cases were CMML with a normal karyotype. The minimum 11q UPD segment is about 2Mb which existed in 11q23. We sequenced coding exons of c-cbl and detected homozygous mutations in 8 of 9 MDS cases with 11q UPD (CMML=5, RAEB=3, overt leukemia=1), but very rare in cases without 11q UPD (1/162), demonstrating that the mutation is tightly linked to 11q UPD. These mutations were 8 point mutations and 1 micro-deletion, they were accumulated in the linker or RING domain. These c-cbl mutants transformed NIH3T3 in a dominant fashion, in which they were phosphorylated and activate PI3K-Akt pathway. To investigate the functions of these mutants in hematopoietic cells, we introduced these mutants into c-kit(+)Sca1(+)Lin(−) murine bone marrow cells, it prolonged replating capacity of these hematopoietic progenitors, suggesting involvement of aberrant c-cbl functions in the myeloproliferative phenotypes frequently found in 11q-UPD positive cases. In conclusion, UPD is an important mechanism of development of MDS, in which both gain-of-function and loss-of-function mutations are duplicated with exclusion of wild-type allele. Analysis of 11q UPD disclosed novel gain-of-function mutations. Identification of the targets of UPDs in 1q, 4q and 7q should also be important to gain a novel insight into the pathogenesis of MDS.