Oncofusion

start / graph / concept-how-a-translocation-forms

Explainerestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.Q1-fusion-genesis

How a chromosome translocation actually forms: proximity + breakage + repair

also called translocation biogenesis · three-step model of translocation · why do chromosomes fuse

For two chromosomes to swap pieces, three things must happen at once: both must snap, the two broken ends must be physically near each other inside the nucleus, and the cell's repair crew must glue the wrong ends together. Remove any one of the three and the fusion never happens.

The same thing, for a scientist

Translocation biogenesis is a three-step process: (1) induction of DNA double-strand breaks (DSBs) at two loci, (2) spatial juxtaposition of the two broken ends, which is heavily constrained by pre-existing non-random 3D genome organisation, and (3) illegitimate end-joining by an error-prone DSB repair pathway. Breakpoint recurrence in cancer is therefore the product of locus-specific fragility and locus-specific spatial proximity, not of a dedicated 'translocase'. This framework is the organising logic for the whole Fusion Genesis beat: every candidate intervention point sits in one of the three steps.

Go deeper
The three-step model in plain words1 deeper

Think of two ropes hanging in a room. To tie the top of rope A to the bottom of rope B you need (a) both ropes cut, (b) the cut ends within arm's reach of each other, and (c) somebody with tape who is careless about which end is which. Cells have all three: nucleases and topoisomerases that cut, a folded nucleus that holds certain genes near each other, and DNA repair pathways that prioritise speed over accuracy 1.

Why this matters for intervention

Each step is a separate, in-principle druggable or detectable event. Step 1 (breakage) is the only step where a preventive small molecule is plausible — and there are candidate targets (TOP2B, APE1, endonuclease G; see mechanism-replication-stress-endog-cleavage-kmt2a). Step 2 (proximity) is set by genome folding and is essentially not druggable without global chromatin disruption. Step 3 (rejoining) is the target of alt-EJ/Polθ inhibitors, but the human genetics argue the wrong pathway is being blocked (see mechanism-nhej-vs-alt-ej-at-fusion-junctions).

Why recurrent partners are recurrent

Because the nucleus is not a bag of spaghetti. Hi-C maps combined with genome-wide translocation sequencing (HTGTS) in mouse pro-B cells showed that, once you normalise the number of DSBs (by irradiating cells so breaks are everywhere), translocation frequency tracks pre-existing spatial proximity — strongly in cis along the same chromosome and within subchromosomal domains 2. Translocation partner choice is therefore substantially a geometry problem 1.

What 'error-prone' means here

The repair pathways that make translocations are the non-homologous ones — they do not need a matching template, they just ligate whatever two ends are held together. That is exactly why the product is a novel junction that exists nowhere in the germline genome, and why every patient's KMT2A::AFF1 junction is at slightly different base pairs. That per-patient uniqueness is clinically exploited for minimal-residual-disease tracking 3.

How this could bite us back
  • Any intervention at step 1 or step 3 hits a pathway the normal cell needs; blocking DSB repair in a fetus is not a therapeutic option, only a mechanistic probe.
Still open
  1. For KMT2A::AFF1 specifically, what is the quantitative split between step-1 fragility and step-2 proximity? Nobody has measured both in the same human fetal haematopoietic progenitor.
  2. Is there any cell state in which all three steps are simultaneously elevated, and is that state transient enough to be an in-utero exposure window?
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review

    Translocation formation requires DSB induction, spatial juxtaposition of broken ends, and error-prone end-joining; non-random genome organisation, chromatin features and chromatin mobility determine translocation frequency and partner selection.

    The biogenesis of chromosome translocations · Roukos V, Misteli T · Nature Cell Biology 2014
    PMID 24691255DOI
    What this evidence cannot show

    Review; synthesises mouse, yeast and human cell-line data. The relative weight of proximity vs fragility differs between systems and is not quantified for any specific human translocation.

    Evidence that argues against this (1)
    • The three requirements are not equally weighted, and the experiment that separates them puts BREAK FREQUENCY well ahead of spatial proximity. Combining a high-resolution Hi-C map of the G1-arrested mouse pro-B cell genome with high-throughput genome-wide translocation sequencing from targeted double-strand breaks, RAG-cleaved antigen-receptor loci were the dominant translocation partners REGARDLESS OF GENOMIC POSITION, reflecting high-frequency breakage at those loci. A proximity contribution only became visible after DSBs were normalised genome-wide with ionising radiation, and then appeared mainly in cis along the broken chromosome and within subchromosomal domains. Citing 'non-random genome organisation determines translocation frequency and partner selection' without that ordering inverts what the data show for a locus that breaks often. Zhang Y et al. 2012
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    After normalising DSB numbers genome-wide by ionising radiation, translocation frequency is directly related to pre-existing spatial proximity measured by Hi-C.

    PMID 22341456DOI
    What this evidence cannot show

    Mouse G1-arrested pro-B cells, engineered DSBs, and irradiation doses far above physiological. Does not directly measure the fetal human progenitor in which KMT2A::AFF1 arises.

    Evidence that argues against this (1)
    • Jasin and colleagues showed that translocation joining mechanism differs between mouse and human cells, so mouse-derived rules for step 3 (and potentially step 2) should not be assumed to transfer. Ghezraoui H et al. 2014
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Genomic breakpoint sequences in KMT2A-rearranged leukaemia are patient-specific and are used clinically to monitor minimal residual disease.

    The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al. · Leukemia 2023
    PMID 37019990DOI
    What this evidence cannot show

    Referral-based cohort of 3401 patients (Frankfurt DCAL); partner-gene frequencies may be biased by referral patterns.

    Evidence that argues against this (1)
    • Patient-specific genomic breakpoints are used for MRD, but the clinical value of that MRD in this disease is more limited than the sentence implies. In 249 infants with KMT2A-rearranged ALL on Interfant-06, MRD was measured by DNA-based PCR of rearranged KMT2A, immunoglobulin and/or T-cell receptor genes at end of induction and end of consolidation. Infants who were MRD-NEGATIVE at end of induction still had a 6-year disease-free survival of only 60.2% (95% CI 43.2-73.6), and the benefit of lymphoid- versus myeloid-style consolidation REVERSED across MRD strata. A patient-specific genomic marker therefore delivers a prognostic gradient, not a reliable statement about cure, and the same breakpoint sequence carries different implications depending on the consolidation given. Stutterheim J et al. 2021
Connected to