Molecular Characterization of the t(15;17)(q22;21) in Epirubicin-Related Acute Promyelocytic Leukaemia

Chromosomal translocations generating chimaeric oncoproteins play an important role in leukaemogenesis, but mechanisms underlying their formation are largely unclear. Substantial insights can be gained from the analysis of therapy-related acute myeloid leukaemias (t-AMLs), which are becoming an increasing healthcare problem as more patients survive their primary cancers. Exposure to agents targeting topoisomerase II (topoII) predisposes to the development of leukaemias with balanced translocations e.g. t(15;17), fusing PML and RARA genes, in therapy-related acute promyelocytic leukaemia (t-APL) which is particularly associated with prior treatment involving mitoxantrone or epirubicin. Using long-range PCR and sequencing to define genomic junction regions we found that in t-APL cases arising in breast cancer patients exposed to the former agent, chromosome 15 breakpoints clustered tightly in an 8 bp “hotspot” region in PML intron 6, which was shown by functional assay to be a preferred site of mitoxantrone-induced DNA topoII cleavage (Mistry et al , NEJM 2005). Subsequent analysis of an independent cohort of t-APL cases arising after mitoxantrone therapy for multiple sclerosis confirmed chromosome 15 breakpoint clustering in the “hotspot” and identified a recurrent breakpoint within RARA intron 2. This hotspot also was shown to be a preferential site of mitoxantrone-induced cleavage in vitro (Hasan et al , Blood 2008). However, the molecular basis of epirubicin-related APL remains uncertain. Therefore we used long-range PCR and sequence analysis to define translocation breakpoints in 6 patients who developed APL after treatment involving epirubicin-containing regimens for prior breast carcinoma. While mitoxantrone-related APL displayed a bias towards breakpoints occurring within PML intron 6 (bcr1), epirubicin-related breakpoints fell within intron 3 (n=2) or intron 6 (n=4) and were outside the mitoxantrone-related “hotspot” (located at position 1484 according to accession number S57791). Breakpoints within the RARA locus were distinct from those observed in mitoxantrone-related APL. Interestingly, the chromosome 15 or chromosome 17 breakpoints of 4 of the epirubicin-related t-APLs fell in close proximity (within 1–4bp) to one of the other cases. Two shared a breakpoint location in PML intron 6 that occurred at bases 1185 and 1186 (accession number S57791) and two fell within RARA intron 2 at base numbers 16193 and 16197 (accession number AJ297538). Given that PML intron 6 and RARA intron 2 are ~1kb and 17kb in length, respectively, such breakpoint clustering was unlikely to have occurred by chance and consistent with functional sites of topoII cleavage. In addition, in vitro DNA cleavage assays demonstrated that heat stable topoII cleavage complexes are formed at the exact location of the breakpoint identified in one of the other patients, at position 1969 in PML intron 6, which were enhanced by the presence of epirubicin. This study suggests that mitoxantrone and epirubicin exhibit preference differences in sites of DNA damage induced by topoisomerase II, which may underlie the propensity to develop specific molecularly-defined subtypes of t-AML according to the nature of the particular chemotherapeutic agent used.