Apoptotic nucleases cut KMT2A, and cells that survive the attempt can carry a translocation
also called apoptotic cleavage MLL bcr · failed apoptosis translocation model · exon 12 apoptotic cleavage site
When a cell starts to kill itself, it chops up its own DNA, and one of the first places it cuts is right inside the KMT2A gene. If the cell then changes its mind and survives, the repair crew can stitch that cut to a cut on another chromosome, leaving a leukaemia-causing fusion behind in a living cell.
The same thing, for a scientist
Apoptotic triggers (ionising radiation, anti-CD95/Fas antibody) induce site-specific cleavage adjacent to KMT2A exon 12 within the BCR, and this cleavage is sufficient to generate KMT2A-restricted translocations in TK6 lymphoblastoid cells that survive. The resulting junctions carry microhomology consistent with non-homologous end joining, and DNA-PKcs localises to the apoptotic cleavage site; wortmannin-mediated suppression of DNA-PKcs increased site-specific cleavage but eliminated KMT2A-restricted translocations, establishing that NHEJ is the joining pathway. Critically, the initial cleavage is TOP2-independent: the catalytic TOP2 inhibitor merbarone did not block it, and a 367 bp episomal fragment of the BCR is sufficient for cleavage at the native position. The model explains a paradox: the same cellular decision that should remove a damaged cell can instead install a founder lesion in a cell that escapes execution.
The escape-from-apoptosis model1 deeper
Betti and colleagues showed that irradiation and anti-CD95 antibody each initiate translocations at the KMT2A exon 12 cleavage site, and proposed that activation of apoptotic effector nucleases alone is sufficient to generate proleukaemogenic translocations, some of which persist in cells that evade apoptotic execution 1. This is the mechanistic reason a stress event with no genotoxin can still produce a fusion, and it fits an in-utero setting where fetal progenitors face waves of physiological apoptosis.
The wortmannin experiment is the key control
Suppressing DNA-PKcs with wortmannin increased site-specific cleavage within KMT2A but eliminated KMT2A-restricted translocations 1. Increased cleavage with abolished translocation cleanly separates step 1 (break) from step 3 (join) and identifies NHEJ as the joiner. It also means blocking the joining step is the only place in this pathway where an intervention would reduce fusion formation, since blocking the break makes things worse.
TOP2-independence
Apoptotic cleavage of the KMT2A BCR occurs at the same position on an episome carrying as little as 367 bp of the BCR, and is not inhibited by merbarone, a catalytic TOP2 inhibitor, so the initial cleavage step is not TOP2-mediated 2. Combined with the de novo vs therapy-related breakpoint split, this supports at least two independent break mechanisms converging on one locus (see breakpoint-kmt2a-bcr-architecture).
Relationship to the EndoG/replication-stress model
These two models are not rivals so much as two triggers of one effector class. EndoG, the nuclease shown to be required for replication-stress-induced MLLbcr cleavage, is itself an apoptotic nuclease that translocates from mitochondria to the nucleus 3. The unresolved question is which upstream trigger dominates in a human fetal progenitor: replication stress, or sub-lethal apoptotic signalling.
| mechanism class | apoptotic nucleolytic cleavage followed by NHEJ |
|---|---|
| cleavage site | adjacent to KMT2A exon 12, within the 8.3 kb BCR |
| minimal sufficient element bp | 367 |
| top2 dependence | independent |
| joining pathway | non-homologous end joining, DNA-PKcs dependent |
triggers
[ "ionising radiation", "anti-CD95/Fas ligation", "sub-lethal apoptotic signalling" ]
- The intervention implied by this model, inhibiting NHEJ, would block the join but leave persistent unrepaired breaks, which is a worse outcome. Wortmannin increased cleavage while blocking translocation only in a cell-line assay where survival was not the endpoint.
- If sub-lethal apoptotic signalling is the trigger, then any therapy that induces incomplete apoptosis in haematopoietic progenitors carries a theoretical risk of installing new fusions, which is exactly the therapy-related leukaemia problem in a new guise.
- Do human fetal liver haematopoietic progenitors undergoing physiological apoptosis show the exon 12 cleavage, and at what frequency?
- If a cell that starts apoptosis and survives is the fusion-founder cell, is there a detectable survivor signature (e.g. residual caspase-cleaved substrates, CAD/ICAD state) that could be used to identify pre-leukaemic clones?
- Nothing here has been replicated outside the Vaughan group in 20 years. Is that because it is wrong, or because nobody tried?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Apoptotic triggers initiate KMT2A-restricted translocations with microhomology-bearing junctions; DNA-PKcs localises to the cleavage site, and wortmannin increases cleavage while eliminating translocations.
Apoptotic triggers initiate translocations within the MLL gene involving the nonhomologous end joining repair system · Betti CJ, Villalobos MJ, Diaz MO, Vaughan AT · Cancer Research 2001What this evidence cannot show
TK6 lymphoblastoid cell line only; wortmannin is a broad PI3K-family inhibitor, so the DNA-PKcs attribution is pharmacological rather than genetic. Translocations were detected by PCR across candidate junctions, which can bias toward detectable products. Single laboratory.
Evidence that argues against this (1)
- Microhomology at the junction does not identify the nuclease, so it cannot be used to argue an apoptotic origin. Genomic junctions from the paradigm prenatal fusion show exactly the same features — microhomologies at the end points plus small deletions and duplications — and the authors explicitly found NO consistent V(D)J signal sequences, topoisomerase II consensus sites or other recombination motifs near the breakpoints, concluding the junctions are compatible with imprecise, CONSTITUTIVE repair of double-strand breaks. Microhomology-bearing junctions are the generic output of error-prone end joining in human cells, whatever made the break, so their presence at KMT2A is consistent with an apoptotic-nuclease origin and equally consistent with several others. Wiemels JL et al. 1999
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Apoptosis-induced cleavage of the KMT2A BCR is TOP2-independent (unaffected by merbarone) and a 367 bp BCR fragment is sufficient for site-specific cleavage on an episome.
Cleavage of the MLL gene by activators of apoptosis is independent of topoisomerase II activity · Betti CJ, Villalobos MJ, Jiang Q, et al. · Leukemia 2005What this evidence cannot show
Episomal (pREP4) context lacks native chromatin; merbarone is not a perfectly selective TOP2 catalytic inhibitor.
Evidence that argues against this (1)
- The etoposide/TOP2B model shows that TOP2 poisoning is certainly sufficient to break the locus, so TOP2-independence of the apoptotic route does not make TOP2 irrelevant, especially for therapy-related disease. Cowell IG et al. 2012
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Endonuclease G, an apoptotic nuclease, accumulates in the nucleus under replication stress and is required for MLLbcr cleavage there, linking the apoptotic-nuclease and replication-stress models through a shared effector.
Endonuclease G initiates DNA rearrangements at the MLL breakpoint cluster upon replication stress. · Gole B, Baumann C, Mian E, Ireno CI, Wiesmuller L · Oncogene 2015What this evidence cannot show
Reporter-based; the connection drawn here between the two literatures is my inference from shared effector identity, not a claim the authors make jointly.
Evidence that argues against this (1)
- Routing KMT2A breaks into microhomology-mediated alternative end joining predicts FEWER translocations, not more, which cuts against using a shared apoptotic/replication-stress effector to explain fusion formation. Polq-null murine cells are selectively hypersensitive to strand-breaking agents, and Polq deficiency produced a MARKED INCREASE in Myc/IgH translocations — the authors state their findings 'depart from the prevailing view that alternative end joining processes are generically translocation-prone' and conclude that POLQ-dependent alt-EJ suppresses genome rearrangements arising at double-strand breaks. A model that supplies more breaks to a suppressive pathway does not by itself supply more fusions; the rate-limiting step may be the failure of that pathway rather than the cleavage. Yousefzadeh MJ et al. 2014
a TOP2/loop-anchor account cannot be complete: apoptotic cleavage of the KMT2A BCR is merbarone-insensitive and episome-sufficient
chlorpyrifos-induced KMT2A rearrangement in fetal liver HSC runs partly through caspase-3 activation
Counter-evidence (Gole & Wiesmuller 2015, PMID 26161385): apoptotic fragmentation, transcription stalling and replication stalling activate nucleases other than TOP2 at MLL, so the TOP2-poison account is one mechanism among several rather than the mechanism
- co-occurs with →Replication stress, AID, base-excision repair and endonuclease G cut the KMT2A locus
two proposed endogenous sources of the double-strand break that starts this disease; they compete as explanations and should be read together
any proposed break mechanism has to explain the position of both cluster regions, not just the major one
- ← related toWhich repair pathway glues the fusion together: classical NHEJ vs alternative end joining
Counter-evidence (Yousefzadeh 2014, PMID 25275444): Polq-defective mice show a MARKED INCREASE in Myc/IgH translocations and the authors state their findings depart from the view that alt-EJ is generically translocation-prone -- POLQ-dependent end joining SUPPRESSES rearrangement, so supplying more breaks to it does not by itself supply more fusions, and Poltheta inhibitors are not prophylaxis
Counter-evidence (Lovett 2001, PMID 11493704): in a therapy-related t(4;11) both sequenced junctions lay at FUNCTIONAL topoisomerase II cleavage sites whose cleavage was enhanced by etoposide and its catechol and quinone metabolites but not dactinomycin -- merbarone-insensitive cleavage in TK6 episomes shows one route to MLL breakage is TOP2-independent, not that the initial cleavage step generally is