Molecular characterization of the t(15;17) in therapy-related APL arising after epirubicin therapy. (A) Distribution of translocation breakpoints within the PML and RARA loci in t-APL cases arising after epirubicin and mitoxantrone. PML exons are represented by red boxes, RARA exons are in blue, and introns are represented by black lines. Arrows indicate the location of PML and RARA translocation breakpoints identified in patients with t-APL arising after mitoxantrone (red arrows) or epirubicin (green arrows), and numbers of the epirubicin-related cases correspond to those presented in Table 1. Details of the mitoxantrone cases have been reported previously.^12,13  (B) PML and RARA breakpoints in epirubicin-related t-APL are preferred sites of epirubicin-induced topoII-mediated DNA cleavage. To identify epirubicin-enhanced cleavage by topoIIα, chromosomal breakpoint junctions were examined in an in vitro assay. DNA cleavage reactions were performed with 25 ng of 5′-labeled DNA (30 000 cpm), 1mM ATP, DMSO, and in the presence or absence of 147nM human DNA topoIIα and 160μM epirubicin. Cleavage complexes were trapped on the addition of SDS and were resolved in an 8% acrylamide–7.0M urea gel. In both panels, reactions in lane 1 were performed with epirubicin (Epi) but lacking DNA topoIIα and show little evidence of cleavage in the absence of the enzyme. Lanes 2 to 5 show dideoxy sequencing reactions primed at the same 5′ end, which allows high-resolution mapping of cleavage sites. Substrates were incubated with topoIIα and DMSO only (lanes 6 and 8) and also in the presence of epirubicin (lanes 7 and 9). Reactions in lanes 8 and 9 were further incubated at 75°C to assess the heat stability of the cleavage complexes. On the left, DNA topoIIα-dependent cleavage is shown within a PML substrate that encompassed the locations of the genomic breakpoints identified in UPN1 and UPN4. The location of the arrows indicate the epirubicin-enhanced heat-stable complexes at position 1184, corresponding precisely to these translocation breakpoints. On the right, cleavage within a substrate that contains the normal homologue of RARA encompassing the breakpoint junction identified in UPN4 is shown, whereby the arrows indicate the epirubicin-enhanced heat-stable complexes corresponding to the der(15) and der(17) translocation breakpoints. (C) Model for formation of the t(15;17) underlying epirubicin-induced t-APL in UPN4. Normal homologues of PML and RARA are indicated in red and blue fonts, respectively. Models show where topoIIα introduces 4-bp staggered nicks in the DNA (as indicated by in vitro experiments), followed by exonucleolytic processing to reveal microhomologies (indicated by gray boxes) that are probably repaired by the error-prone nonhomologous end joining repair pathway. Template-directed polymerization (indicated with black font), mismatch repair (represented by green font), and ligation fills in any remaining gaps to generate the PML-RARA and RARA-PML genomic breakpoint junctions that were identified in the t-APL arising in this patient.