A distinct population of therapy-related acute myeloid leukemia (t-AML) is strongly associated with prior administration of topoisomerase II (topo II) inhibitors. These t-AMLs display distinct cytogenetic alterations, most often disrupting the MLL gene on chromosome 11q23 within a breakpoint cluster region (bcr) of 8.3 kb. We recently identified a unique site within the MLL bcr that is highly susceptible to DNA double-strand cleavage by classic topo II inhibitors (e.g., etoposide and doxorubicin). Here, we report that site-specific cleavage within the MLL bcr can be induced by either catalytic topo II inhibitors, genotoxic chemotherapeutic agents which do not target topo II, or nongenotoxic stimuli of apoptotic cell death, suggesting that this site-specific cleavage is part of a generalized cellular response to an apoptotic stimulus. We also show that site-specific cleavage within the MLL bcr can be linked to the higher-order chromatin fragmentation that occurs during the initial stages of apoptosis, possibly through cleavage of DNA loops at their anchorage sites to the nuclear matrix. In addition, we show that site-specific cleavage is conserved between species, as specific DNA cleavage can also be demonstrated within the murine MLL locus. Lastly, site-specific cleavage during apoptosis can also be identified at the AML1 locus, a locus which is also frequently involved in chromosomal rearrangements present in t-AML patients. In conclusion, these results suggest the potential involvement of higher-order chromatin fragmentation which occurs as a part of a generalized apoptotic response in a mechanism leading to chromosomal translocation of the MLL and AML1 genes and subsequent t-AML.Nonrandom chromosomal aberrations, particularly chromosomal translocations, are frequently found in association with a wide spectrum of malignancies, most prominently leukemias and lymphomas (9,14,47). The available evidence suggests that these nonrandom chromosomal translocations are often causal events leading to malignant transformation (2, 40). However, the molecular mechanisms which cause these translocations remain largely unknown.In many cases, a powerful argument can be made that these translocations are the result of mistakes in normal V(D)J recombinase activity (4, 60). These arguments are based on identification of features that resemble normal V(D)J recombinase activity, such as cryptic heptamer sequences, nontemplated N-region nucleotide addition, and exonucleolytic deletion of germ line nucleotides at the translocation breakpoints (4, 60). Other factors that have been implicated in the generation of nonrandom translocations include homologous recombination events between Alu elements (52) and exposure to DNA-damaging agents (35,36). It has been recognized for some time that DNA-damaging cancer chemotherapeutic agents, such as the topoisomerase II (topo II) inhibitor etoposide (36) and the alkylating agent melphalan (35), can cause chromosomal translocations. For instance, phytohemagglutinin-stimulated peripheral blood ly...

The t(12;21)(p13;q22) translocation, fusing the ETV6 andAML1 genes, is the most frequent chromosomal translocation associated with pediatric B-cell precursor acute lymphoblastic leukemia. Although the genomic organization of the ETV6 gene and a breakpoint cluster region (bcr) in ETV6 intron 5 has been described, mapping of AML1 breakpoints has been hampered because of the large, hitherto unknown size of AML1 intron 1. Here, we report the mapping of the AML1 gene between exons 1 and 3, cloning of ETV6-AML1 breakpoints from different patients, and localization of the AML1 breakpoints withinAML1 intron 1. In contrast to the tightly clustered ETV6breakpoints, the AML1 breakpoints were found to be dispersed throughout AML1 intron 1. Although nucleotide sequence analysis of the breakpoint junctions showed several 5/7 matches for the V(D)J consensus heptamer recognition sequence, these matches were present only on the ETV6 alleles and not on the AML1 alleles, making it unlikely that the translocations were mediated by a simple V(D)J recombination mistake. Interestingly, several breakpoints as well as a stable insertion polymorphism mapped close to a polymorphic, alternating purine-pyrimidine tract in the ETV6 gene, suggesting that this region may be prone to DNA recombination events such as insertions or translocations. Finally, the presence of an insertional polymorphism within the ETV6 bcr must be recognized to avoid incorrect genotype designation based on Southern blot analysis.

Aims-The MLL gene on chromosome 11q23 is frequently disrupted by chromosomal translocations in association with haematological malignancies. Recently, a specific site within the 8.3 kb MLL breakpoint cluster region that is cleaved during the early stages of apoptosis has been identified. Because MLL gene rearrangements are used to identify patients with high risk leukaemia, it was the aim of this study to determine whether this DNA cleavage event could be triggered in diagnostic bone marrow samples solely through ex vivo incubation at room temperature. Methods-Pretreatment bone marrow samples were collected from six paediatric leukaemia patients. Genomic DNA for Southern blot analysis of MLL gene rearrangements was isolated immediately after samples were obtained and compared to genomic DNA isolated after incubation of specimens for 24-60 hours at room temperature, simulating delays in processing that might occur when samples are delivered to reference laboratories. In addition, cryopreserved samples from 70 paediatric leukaemia patients were screened for evidence of site specific MLL cleavage. Results-After ex vivo incubation of bone marrow samples, site specific MLL cleavage resulting in a pseudo-rearrangement of the MLL gene was detected in two of six patients. In addition, a third patient with a similar MLL pseudo-rearrangement in cryopreserved cells was identified. Conclusion-Pseudo-rearrangement of the MLL gene at chromosome 11q23 was caused by ex vivo incubation of bone marrow samples. This novel phenomenon, which could lead to misclassification of leukaemia patients, might also be of importance for genotype analysis by Southern blotting at other loci. (J Clin Pathol: Mol Pathol 1998;51:85-89)

Figure 2IgH and Ig rearrangement analysis at diagnosis and relapses. Genomic DNA from diagnosis and relapses was subjected to PCR analysis using the FR1c-5Ј and JHc-3Ј IgH-specific primers and using the IRSS-5Ј and Kde-3Ј Ig-specific primers. Both IgH and Igspecific PCR products were spotted in duplicate on to the nylon membrane and hybridized with the patient-specific probe. Mononuclear cells (MNC) from a healthy volunteer donor served as a control. Dotblot results for IgH and Ig are shown in a and b, respectively. because the same IgH chain rearrangement was already present at the onset, while chain gene deletion was detectable at relapse only.This case underlines the important role of comparative studies, mainly at the molecular level, between disease patterns at diagnosis and relapse, as potential means yielding valuable insights into the mechanism of leukemic recurrence. Therapy-related T cell lymphoblastic lymphoma with t(11;19)(q23;p13) and MLL gene rearrangement TO THE EDITORDuring the past several decades, advances in the treatment of childhood cancers have resulted in increased survival of pediatric cancer patients. This is in large part due to the use of aggressive chemotherapy regimens, radiation therapy or a combination of both. However, along with the benefits of increased survival. It has become Correspondence: PD Aplan at his current address: