Translocation breakpoints involving the MLL gene linked to Infant Acute Leukemia (IAL) and therapy related acute leukemia (tAL) are tightly clustered between MLL exons 8 and 12. Exon 12 also marks the location of a well-described cleavage hotspot that is synchronous with a sharp decline in total MLL fusions observed in clinical samples. Though multiple MLL fusion partners have been identified, fusions to MLLT3 (AF9) and AFF1 (AF4) comprise 56% of all clinical rearrangements so far assayed. Epidemiological data has linked maternal exposure to birth control formulations with an increased risk of IAL involving MLL gene rearrangements. Subsequent in-vitro studies suggested a role of estradiol (E2) in the generation of such rearrangements. In order to probe the action of E2 in generating these lesions, the ability of E2 to impact MLL rearrangement formation was studied, focusing on the exon 12 hotspot and using immortalized but not transformed TK6 lymphoblastoid cells. Real-time PCR studies showed that in this cell line, transient exposure to 10 nM E2 enhanced transcription of MLL eight fold over controls. E2 treatment also increased transcription of MLL partner genes, MLLT3 and AFF1, through to a lesser degree. To determine if the process of transcription led to gene co-localization, chromatin conformation capture (3C) experiments were performed. Here, brief exposure to 10 nM E2 led to the co-localization of MLL with MLLT3, using primer sets targeting both MLL introns 9 and 13 and MLLT3 introns 4 and 8. These data indicated contact between these two genes, over a substantial region, consistent with their occupation of an operationally defined “transcription factory”. Surprisingly, low levels of E2 also stimulated the generation of de-novo MLL- MLLT3 fusion transcripts, without the application of any genotoxic stressors. To identify the process whereby each gene is fragmented, an essential precursor to any rearrangement, RNAi knockdown of activation induced cytidine deaminase (AICDA) was studied. AICDA activity is normally associated with class switch recombination and somatic hypermutation, but has recently been identified as the fragmenting agent associated with c-myc/IgH fusions in Burkitts lymphoma. RNAi knockdown of AICDA suppressed the induction of MLL-MLLT3 fusion transcript formation observed with E2 treatment, suggesting AICDA involvement in MLL and partner gene fragmentation. To probe this association in more detail, a ChIP analysis was performed targeting AICDA recruitment to MLL intron 11 (hot spot for rearrangements) or intron 12 (few MLL rearrangements). Here, E2 dependent localization of AICDA was noted upstream of the DNA cleavage hotspot and within the region of elevated MLL fusions in intron 11, but not in a region showing few rearrangements. Combined, these studies show that concentrations of E2 that occur during pregnancy, or during use of oral contraceptives, have the potential to initiate MLL fusions through an endogenous AICDA-mediated mechanism, that is enhanced by gene proximity associated with synchronous transcription of both MLL and partner genes. Further, the link between transcription-induced co-localization and MLL rearrangements may identify a common mechanism of MLL fusion gene formation relevant to a wider range of clinical diagnoses. If correct, then this mechanism may be a target for manipulation, particularly in controlled settings such as the delivery of potentially leukemogenic therapeutics.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution