Rearrangements involving the MLL gene on chromosome band 11q23 are a hallmark of therapy-related acute myeloid leukemias following treatment with topoisomerase II poisons including etoposide. Therapy-related and de novo genomic translocation breakpoints cluster within a well-characterized 8.3-kb fragment of MLL. Repair of etoposidestabilized DNA topoisomerase II covalent complexes may initiate MLL rearrangements observed in patients. We used a culture system of primary human hematopoietic CD34 ؉ cells and inverse polymerase chain reaction to characterize the spectrum of stable genomic rearrangements promoted by etoposide exposure originating within an MLL translocation hotspot in therapy-related leukemia. Alterations to the region were observed at a readily detectable frequency in etoposidetreated cells. Illegitimate repair events after minimal repair included MLL tandem duplications and translocations, with minor populations of deletions or insertions. In stably repaired cells that proliferated for 10 to 14 days, the significant majority of illegitimate events were MLL IntroductionGenome integrity is controlled by multiple damage surveillance pathways, cell-cycle checkpoints, and repair mechanisms. [1][2][3] Despite these safeguards, illegitimate repair of chromosomal doublestrand breaks (DSBs), such as those produced by chemotherapy agents that target topoisomerase II (topo II), can promote chromosomal rearrangements and, ultimately, tumorigenesis. Topo II is an essential cellular enzyme that catalyzes changes in DNA topology via its cleavage-religation equilibrium. Topo II-targeted drugs that are poisons of this enzyme stabilize topo II-DNA covalent complexes, most often by decreasing the rate of religation in a dose-dependent manner. Disruption of the cleavage-religation reaction results in accumulation of DSBs, p53 activation, and induction of apoptosis or repair. 4,5 Chromosomal DSBs are potent inducers of recombination, stimulating the exchange of homologous sequences between 2 DNA duplexes 1000-fold 6-8 not only between sister chromatids, but also between homologs, and sequence repeats on heterologous chromosomes. 9-12 DSBs may be repaired by homologous recombination or nonhomologous end joining (NHEJ), and both repair mechanisms have been associated with chromosomal translocations, a hallmark of leukemias, lymphomas, and soft-tissue sarcomas. [13][14][15] There is clear evidence that exposure to chemotherapy or irradiation can result in subsequent development of therapy-related leukemias, which occur in 1% to 15% of patients exposed to DNA-damaging agents in anticancer regimens. 16,17 Exposure to topo II poisons such as etoposide is predominantly associated with therapy-related leukemias characterized by rearrangements of the MLL gene on chromosome band 11q23. 16 MLL spans 100 kb, encodes a 430-kDa protein homologous to the Drosophila trithorax gene, and has important functions in embryogenesis and hematopoiesis. 18,19 Mixed-lineage leukemia (MLL) is a critical transcriptional regulator and numerous tr...

An inactivating polymorphism at position 609 in the NAD(P)H:quinone oxidoreductase 1 gene (NQO1 C609T) is associated with an increased risk of adult leukemia. A small British study suggested that NQO1 C609T was associated with an increased risk of infant leukemias with MLL translocations, especially infant acute lymphoblastic leukemia (ALL) with t(4; 11

Few t(9;11) translocations in DNA topoisomerase II inhibitor-related leukemias have been studied in detail and the DNA damage mechanism remains controversial. We characterized the der(11) and der(9) genomic breakpoint junctions in a case of AML following etoposide and doxorubicin. Etoposide-, etoposide metabolite- and doxorubicin-induced DNA topoisomerase II cleavage was examined in normal homologues of the MLL and AF-9 breakpoint sequences using an in vitro assay. Induction of DNA topoisomerase II cleavage complexes in CEM and K562 cell lines was investigated using an in vivo complex of enzyme assay. The translocation occurred between identical 5'-TATTA-3' sequences in MLL intron 8 and AF-9 intron 5 without the gain or loss of bases. The 5'-TATTA-3' sequences were reciprocally cleaved by DNA topoisomerase II in the presence of etoposide, etoposide catechol or etoposide quinone, creating homologous 4-base 5' overhangs that would anneal to form both breakpoint junctions without any processing. der(11) and der(4) translocation breakpoints in a treatment-related ALL at the same site in MLL are consistent with a damage hotspot. Etoposide and both etoposide metabolites induced DNA topoisomerase II cleavage complexes in the hematopoietic cell lines. These results favor the model in which the chromosomal breakage leading to MLL translocations in DNA topoisomerase II inhibitor-related leukemias is a consequence of DNA topoisomerase II cleavage.

We used panhandle PCR to clone the der(11) genomic breakpoint junction in three leukemias with t(4;11) and devised reversepanhandle PCR to clone the breakpoint junction of the other derivative chromosome. This work contributes two elements to knowledge on MLL translocations. First is reverse-panhandle PCR for cloning breakpoint junctions of the other derivative chromosomes, sequences of which are germane to understanding the MLL translocation process. The technique revealed duplicated sequences in one case of infant acute lymphoblastic leukemia (ALL) and small deletions in a case of treatment-related ALL. The second element is discovery of a three-way rearrangement of MLL, AF-4, and CDK6 in another case of infant ALL. Cytogenetic analysis was unsuccessful at diagnosis, but suggested t(4;11) and del(7)(q21q31) at relapse. Panhandle PCR analysis of the diagnostic marrow identified a breakpoint junction of MLL intron 8 and AF-4 intron 3. Reverse-panhandle PCR identified a breakpoint junction of CDK6 from band 7q21-q22 and MLL intron 9. CDK6 encodes a critical cell cycle regulator and is the first gene of this type disrupted by MLL translocation. Cdk6 is overexpressed or disrupted by translocation in many cancers. The in-frame CDK6-MLL transcript is provocative with respect to a potential contribution of the predicted Cdk6-MLL fusion protein in the genesis of the ALL, which also contains an in-frame MLL-AF4 transcript. The sequences in these three cases show additional MLL genomic breakpoint heterogeneity. Each breakpoint junction suggests nonhomologous end joining and is consistent with DNA damage and repair. CDK6-MLL is a new fusion of both genes.

We examined the MLL translocation in two cases of infant AML with X chromosome disruption. The Gbanded karyotype in the first case suggested t(X;3)(q22;p21)ins(X;11)(q22;q13q25). Southern blot analysis showed one MLL rearrangement. Panhandle PCR approaches were used to identify the MLL fusion transcript and MLL genomic breakpoint junction. SEPTIN6 from chromosome band Xq24 was the partner gene of MLL. MLL exon 7 was joined in-frame to SEPTIN6 exon 2 in the fusion transcript. The MLL genomic breakpoint was in intron 7; the SEPTIN6 genomic breakpoint was in intron 1. Spectral karyotyping revealed a complex rearrangement disrupting band 11q23. FISH with a probe for MLL confirmed MLL involvement and showed that the MLL-SEPTIN6 junction was on the der(X). The MLL genomic breakpoint was a functional DNA topoisomerase II cleavage site in an in vitro assay. In the second case, the karyotype revealed t(X;11)(q22;q23). Southern blot analysis showed two MLL rearrangements. cDNA panhandle PCR detected a transcript fusing MLL exon 8 in-frame to SEPTIN6 exon 2. MLL and SEPTIN6 are vulnerable to damage to form recurrent translocations in infant AML. Identification of SEPTIN6 and the SEPTIN family members hCDCrel and MSF as partner genes of MLL suggests a common pathway to leukaemogenesis.

No abstract

Suggested trends warranting investigation in more patients were breakpoint sites in the 3' bcr in AML and in patients older than 12 months. The distribution of MLL genomic breakpoints within the bcr in de novo leukemia in children is distinct from that in adults, where the breakpoints cluster in the 5' portion of the bcr.

Identifying translocations of the MLL gene at chromosome band 11q23 is important for the characterization and treatment of leukemia. However, cytogenetic analysis does not always find the translocations and the many partner genes of MLL make molecular detection difficult. We developed cDNA panhandle PCR to identify der(11) transcripts regardless of the partner gene. By reverse transcribing first-strand cDNAs with oligonucleotides containing coding sequence from the 5 MLL breakpoint cluster region at the 5 ends and random hexamers at the 3 ends, known MLL sequence was attached to the unknown partner sequence. This enabled the formation of stem-loop templates with the fusion point of the chimeric transcript in the loop and the use of MLL primers in two-sided PCR. The assay was validated by detection of the known fusion transcript and the transcript from the normal MLL allele in the cell line MV4 -11. cDNA panhandle PCR then was used to identify the fusion transcripts in two cases of treatment-related acute myeloid leukemia where the karyotypes were normal and the partner genes unknown. cDNA panhandle PCR revealed a fusion of MLL with AF-10 in one case and a fusion of MLL with ELL in the other. Alternatively spliced transcripts and exon scrambling were detectable by the method. Leukemias with normal karyotypes may contain cryptic translocations of MLL with a variety of partner genes. cDNA panhandle PCR is useful for identifying MLL translocations and determining unknown partner sequences in the fusion transcripts.T ranslocations of the MLL gene at chromosome band 11q23 occur in leukemias of infants (reviewed in ref. 1) and leukemias associated with DNA topoisomerase II inhibitors (reviewed in ref.2). The ability to rapidly identify MLL translocations, whether by cytogenetic or molecular approaches, is relevant to diagnosis and prognosis and to treatment planning. MLL is an example of a gene involved in translocations with numerous different partner genes; many are still uncharacterized (1, 2). The specific partner gene with which MLL is fused may have an impact on the clinical response (3). Previously, we developed panhandle PCR approaches to identify MLL der(11) translocation breakpoints in genomic DNA (4-8). The salient features include attachment of known MLL sequence to the unknown partner gene, formation of a stem-loop template, and two-sided PCR. Because MLL sequences are at both ends of the template, all primers are derived from MLL. Panhandle PCR methods offer the advantage of amplifying MLL translocation breakpoints without primers from the partner genes.Although panhandle PCR approaches are highly effective for genomic translocation breakpoint cloning (4-8), sometimes the genomic target sequence may be too large to amplify. In addition, if the amplicon contains intronic sequence only and not exonic sequence, panhandle PCR approaches may not reveal the partner gene (4,8). MLL genomic translocation breakpoints occur within an 8.3-kb breakpoint cluster region (bcr) (9). Here, we targeted the corresponding 85...