The t(7;11)(p15;p15) translocation is a recurrent chromosomal abnormality associated primarily with acute myeloid leukaemia (FAB M2 and M4). We present here the molecular definition of this translocation. On chromosome 7 positional cloning revealed the consistent rearrangement of the HOXA9 gene, which encodes a class I homeodomain protein potentially involved in myeloid differentiation. On chromosome 11 the translocation targets the human homologue of NUP98, a member of the GLFG nucleoporin family. Chimaeric messages spliced over the breakpoint fuse the GLFG repeat domains of NUP98 in-frame to the HOXA9 homeobox. The predicted NUP98-HOXA9 fusion protein may promote leukaemogenesis through inhibition of HOXA9-mediated terminal differentiation and/or aberrant nucleocytoplasmic transport.

Chronic myeloid leukaemia (CML) develops when two genes, BCR on chromosome 22 and ABL on chromosome 9, recombine to form a hybrid BCR-ABL gene with leukaemogenic properties. The mechanism which underlies this recombination is unknown, but additional chromosome sites may be involved to form complex BCR-ABL rearrangements. The majority of breakpoints in BCR occur within a 5 kb major breakpoint cluster region, M-Bcr. Here, we show that the 3' part of M-Bcr recombined within, or immediately adjacent to, Alu elements at the additional sites in all five complex BCR-ABL rearrangements that have been examined so far. This is a new finding which suggests that Alu sequences have an affinity for the BCR-ABL recombination process in complex rearrangements, and provides additional evidence for the association of these elements with somatic rearrangements which cause human leukaemia. We further show that sequence motifs similar to IgH switch pentamers and consensus binding sites of the lymphoid-associated Translin protein are present on one or more participating strands at 3'M-Bcr recombination sites. Motifs similar to Translin-binding sites were also identified within the Alu consensus. Expressed sequences mapped close to the breakpoint sites on other chromosomes in three of the five cases examined.

Genomic rearrangements are a well-recognized cause of genetic disease and can be formed by a variety of mechanisms. We report a complex rearrangement causing severe hemophilia A, identified and further characterized using a range of PCR-based methods, and confirmed using array-comparative genomic hybridization (array-CGH). This rearrangement consists of a 15.5-kb deletion/16-bp insertion located 0.6 kb from a 28.1-kb deletion/263-kb insertion at Xq28 and is one of the most complex rearrangements described at a DNA sequence level. We propose that the rearrangement was generated by distinct but linked cellular responses to double strand breakage, namely break-induced replication (BIR) and a novel model of break-induced serial replication slippage (SRS). The copy number of several genes is affected by this rearrangement, with deletion of part of the Factor VIII gene (F8, causing hemophilia A) and the FUNDC2 gene, and duplication of the TMEM185A, HSFX1, MAGEA9, and MAGEA11 genes. As the patient exhibits no clinically detectable phenotype other than hemophilia A, it appears that the biological effects of the other genes involved are not dosage-dependent. This investigation has provided novel insights into processes of DNA repair including BIR and the first description of SRS during repair in a pathological context.

RUNX1 (AML1, CBFA2) is mutated in affected members of families with autosomal dominant thrombocytopenia and platelet dense granule storage pool deficiency. Many of those affected, usually by point mutations in one allele, are predisposed to the development of acute myeloid leukemia (AML) in adult life. The RUNX1 protein complexes with core binding factor beta (CBFB) to form a heterodimeric core binding transcription factor (CBF) that regulates many genes important in hematopoiesis. RUNX1 was first identified as the gene on chromosome 21 that is rearranged by the translocation t(8;21)(q22;q22.12) recurrently found in the leukemic cells of patients with AML. In addition to the t(8;21), RUNX1 is rearranged with one of several partner genes on other chromosomes by somatically acquired translocations associated with hematological malignancies. Point mutations of RUNX1 are also found in sporadic leukemias to reinforce the important position of this gene on the multi-step path to leukemia. In animal models, at least one functional copy of RUNX1 is required to effect definitive embryonic hematopoiesis. Cells expressing dominant-negative mutants of RUNX1 are readily immortalized and transformed, and those RUNX1 mutants which retain CBFB binding ability may possess dominant-negative function. However, in some families there is transmitted one mutated allele of RUNX1 with no dominant-negative function, demonstrating that simple haploinsufficiency of RUNX1 predisposes to AML and also causes a generalized hematopoietic stem cell disorder most recognizable as thrombocytopenia.

Summary. The RUNX1 (AML1, CBFA2) gene is a member of the runt transcription factor family, responsible for DNA binding and heterodimerization of other non-DNA binding transcription factors. RUNX1 plays an important part in regulating haematopoiesis and it is frequently disrupted by illegitimate somatic recombination in both acute myeloid and lymphoblastic leukaemia. Germline mutations of RUNX1 have also recently been described and are dominantly associated with inherited leukaemic conditions. We have identified a unique point mutation of the RUNX1 gene (A107P) in members of a family with autosomal dominant inheritance of thrombocytopenia. One member has developed acute myeloid leukaemia (AML).