Chromatin loop anchors are DNA-break hotspots - and the two papers that established it disagree about why
also called CTCF/cohesin loop anchor fragility · TOP2B breakage at loop bases · loop-anchor fragile sites
The genome is folded into loops, and the places where a loop is pinned shut turn out to be where the DNA most often snaps. Two teams showed this within two years of each other and reached opposite conclusions about whether the snapping needs the gene to be switched on.
The same thing, for a scientist
Two 2017-2019 studies established that CTCF/cohesin-bound chromatin loop anchors accumulate topoisomerase-II-mediated double-strand breaks and that recurrently translocated cancer genes are enriched there. They disagree on the role of transcription. Canela et al. (2017) mapped DSBs by END-seq and reported that loop-anchor breakage is TOP2B-dependent but largely transcription-, replication- and cell-type-INDEPENDENT, that polymorphisms redistributing CTCF/cohesin occupancy rewire the cleavage sites to novel anchors, and that DSBs are enriched on both sides of strong TAD borders and fall at breakpoint clusters commonly translocated in cancer [E1]. Gothe et al. (2019) mapped TOP2 DSBs and reported the opposite emphasis: enrichment at loop anchors with HIGH transcriptional activity, break frequency positively correlated with transcriptional output and directionality, and MLL/KMT2A fusion formation at recurrent hotspots dependent on both TOP2 isoforms AND on transcriptional elongation [E2]. Canela et al. (2019) then partially reconciled them by separating two steps: TOP2 chromatin localisation and etoposide trapping are transcription-independent but require pre-existing cohesin binding, whereas CONVERSION of a trapped cleavage complex into an irreversible DSB is accelerated about two-fold at transcribed loci and depends on proteasomal degradation and TDP2 phosphodiesterase activity [E3]. The synthesis that survives all three is: cohesin puts the enzyme there; transcription decides how often the lesion becomes permanent.
What a loop anchor is, and why it would break1 deeper
Cohesin extrudes chromatin until it is blocked by convergently oriented CTCF sites; the blocked positions are loop anchors and TAD borders. TOP2 relieves the torsional stress that extrusion and transcription both generate, and it does so by cutting both strands and holding the ends covalently in a cleavage complex (TOP2cc). A place where the enzyme is concentrated and the DNA is held taut is a place where an abortive reaction leaves a break. Canela et al. showed this is not a correlation artefact: polymorphisms that move CTCF/cohesin occupancy move the cleavage sites with them, to novel anchors 1. That is as close to a causal test of the geometry as this literature has.
The measurement instrument matters
END-seq monitors DNA ends genome-wide at base-pair resolution and can detect at least one DSB per cell among 10,000 cells that do not have one 4. That sensitivity is what makes a claim about endogenous, continuously forming breaks possible at all; the older literature could only see breaks after a genotoxic dose. It also sets the limit: END-seq reports ends present in a population at a moment, so a lesion that is rapidly and correctly religated is under-counted relative to one that persists.
The disagreement, stated plainly
Canela 2017: loop-anchor DSBs are largely transcription-, replication- and cell-type-independent and form continuously throughout interphase 1. Gothe 2019: loop-anchor DSBs are enriched where transcription is high, scale with transcriptional output and directionality, and the resulting MLL fusions require transcriptional elongation 2. These are not compatible as stated, and this graph should not cite one as if the other did not exist.
The partial reconciliation is Canela 2019's two-step decomposition 3. Trapping is cohesin-dependent and transcription-independent; conversion to an irreversible break is transcription-accelerated and TDP2/proteasome-dependent. Under that model Canela 2017 measured a quantity dominated by step one and Gothe measured a quantity dominated by step two, and both are right about what they measured. This reconciliation is the entry's own reading, not a claim made by any of the three papers, and it is offered as the most economical account rather than as a settled result.
Does the KMT2A breakpoint cluster region sit at a loop anchor?
Not established at the resolution the question deserves. Gothe et al. state that genes recurrently translocating to drive leukaemias are highly transcribed and enriched at loop anchors, and that these genes accumulate DSBs at recurrent hotspots giving rise to chromosomal fusions 2. Canela et al. state that loop-anchor DSBs frequently occur at breakpoint clusters commonly translocated in cancer 1. Both are gene-level or class-level statements.
Neither paper is cited here as showing that the 8.3 kb KMT2A BCR itself coincides with a specific CTCF/cohesin anchor in a haematopoietic progenitor, and the graph's breakpoint-kmt2a-bcr-architecture asks exactly that question of the intron 11 / exon 12 hotspot. Answering it requires overlaying a progenitor CTCF/cohesin ChIP and loop call on the BCR coordinates; the analysis is specified in ana-repeat-feature-enrichment-kmt2a-breakpoints.
| mechanism class | 3D genome architecture / topoisomerase-mediated DNA breakage |
|---|---|
| structural feature | CTCF/cohesin-bound chromatin loop anchor and TAD border |
| unresolved disagreement | Canela 2017 (E1) reports loop-anchor breakage as transcription-, replication- and cell-type-independent; Gothe 2019 (E2) reports it as transcription-output-dependent. Unreconciled in the primary literature. |
| kmt2a specific status | NOT established that the 8.3 kb KMT2A BCR coincides with a named CTCF/cohesin loop anchor in any haematopoietic progenitor. Gene-level and class-level enrichment only. |
effectors
[ "TOP2B", "TOP2A", "CTCF", "cohesin (RAD21/SMC1A/SMC3)", "TDP2", "26S proteasome" ]
two step model
{
"step_1_trapping": "TOP2 localisation and etoposide trapping: cohesin-dependent, transcription-independent",
"step_2_conversion": "trapped TOP2cc to irreversible DSB: about 2-fold faster at transcribed loci, proteasome- and TDP2-dependent",
"source": "E3"
}assays
[ "END-seq (E4)", "sBLISS (used in E2)", "Hi-C and loop calling", "CTCF and cohesin ChIP-seq" ]
- If loop-anchor fragility is a generic architectural property of the genome, then anything that reduced it - cohesin or CTCF perturbation - would deform the whole genome's regulatory architecture. This is not a druggable prevention target; it is a description of how genomes are built.
- Transcription inhibition would reduce the conversion step [E3] and is exactly the intervention a fetal haematopoietic progenitor can least afford.
- TDP2 loss reduces one rare junction class but INCREASES overall TOP2-induced translocation (see `mechanism-tdp2-repair-of-trapped-top2-complexes`), so blocking the conversion machinery makes the outcome worse, not better.
- Does the KMT2A BCR, and specifically the intron 11 / exon 12 hotspot, overlap a called CTCF/cohesin loop anchor in human fetal liver HSC/LMPP or cord-blood CD34+ cells? This is a public-data overlay and nobody appears to have published it.
- Canela 2017 says loop-anchor breakage is cell-type-independent. If that holds at the KMT2A BCR then the cell-type specificity of KMT2A-r leukaemia is entirely downstream of the break and none of it is breakage geography. That is a strong, testable claim and it has never been tested at this locus.
- Does a cohesin (RAD21) degron change the position of etoposide-induced breaks WITHIN the KMT2A BCR, as the polymorphism-rewiring result [E1] predicts it should?
- The two-step model [E3] predicts that a transcription elongation inhibitor should reduce irreversible DSB formation at the KMT2A BCR without reducing TOP2 trapping there. Nobody has run that experiment at this locus.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
CTCF/cohesin-bound chromosome loop anchors are vulnerable to TOP2B-mediated DSBs; polymorphisms redistributing CTCF/cohesin occupancy rewire the cleavage sites to novel anchors, and loop-anchor DSBs are largely transcription-, replication- and cell-type-independent and fall at breakpoint clusters commonly translocated in cancer.
Genome Organization Drives Chromosome Fragility · Canela A, Maman Y, Jung S, et al. · Cell 2017What this evidence cannot show
Mouse and human transformed lines; END-seq measures ends present in a population, so rapidly religated cleavage complexes are undercounted. 'Frequently occur at breakpoint clusters commonly translocated in cancer' is a genome-wide association statement and is not a measurement at the KMT2A BCR. Read from the abstract; the full text is in PMC (PMC6133249) but is not open access.
Evidence that argues against this (1)
- The transcription-independence claim is directly contradicted by the other founding paper of this model, which reports that TOP2-mediated DSBs are enriched at loop anchors with HIGH transcriptional activity, that DSB frequency positively correlates with transcriptional output and directionality, and that the resulting oncogenic MLL fusions depend on transcriptional elongation. Any entry citing loop-anchor fragility as a transcription-independent property of the genome is citing one half of an unresolved disagreement. Gothe HJ et al. 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
TOP2-mediated DSBs are enriched at chromatin loop anchors with high transcriptional activity, recur at CTCF- and cohesin-bound loop bases, correlate positively with transcriptional output and directionality, and give rise to oncogenic MLL fusions in a manner dependent on both TOP2 isoforms and on transcriptional elongation.
Spatial Chromosome Folding and Active Transcription Drive DNA Fragility and Formation of Oncogenic MLL Translocations · Gothe HJ, Bouwman BAM, Gusmao EG, et al. · Molecular cell 2019What this evidence cannot show
Read from the abstract only. Route 1 of the four-route unreadable test returned NO PMCID for this PMID (Europe PMC core record checked 2026-09-20: pmcid null, inEPMC N), so routes 2-4 are unavailable and the full text was not read. The fusion-formation experiments are in cell lines under etoposide, not in a fetal progenitor and not without drug.
Evidence that argues against this (1)
- The same year, the group that established loop-anchor fragility reported that TOP2 chromatin localisation and etoposide trapping are INDEPENDENT of transcription and instead require pre-existing cohesin binding, with transcription affecting only the downstream conversion step. So 'active transcription drives DNA fragility' is an overstatement of a two-step process whose first step is transcription-independent. Canela A et al. 2019
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Etoposide trapping of TOP2 cleavage complexes is transcription-independent but requires pre-existing cohesin binding, whereas conversion of trapped complexes into irreversible TOP2-linked DSBs is accelerated about two-fold at transcribed loci and depends on proteasomal degradation and TDP2 phosphodiesterase activity; cohesin binding and transcriptional activity alone predict the kinetics of TOP2-induced DSBs.
Topoisomerase II-Induced Chromosome Breakage and Translocation Is Determined by Chromosome Architecture and Transcriptional Activity · Canela A, Maman Y, Huang SN, et al. · Molecular cell 2019What this evidence cannot show
The whole decomposition is measured under exogenous etoposide. It is a model of how a TOP2 poison produces breaks, and its transfer to spontaneous, drug-free breakage in a fetal progenitor is an assumption, not a result. Read from the abstract.
Evidence that argues against this (1)
- The two-step model is derived entirely from drug-trapped complexes, and the KMT2A locus has a documented route to cleavage that does not involve TOP2 at all: apoptosis-induced cleavage within the KMT2A BCR is unaffected by the catalytic TOP2 inhibitor merbarone, and a 367 bp episomal fragment of the BCR is sufficient for site-specific cleavage. A loop-anchor/TOP2 account of KMT2A fragility therefore cannot be the whole mechanism even in principle. Betti CJ et al. 2005
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
END-seq monitors DNA double-strand breaks and end resection genome-wide at base-pair resolution in vivo and can detect at least one DSB per cell among 10,000 cells not harbouring DSBs.
DNA Breaks and End Resection Measured Genome-wide by End Sequencing · Canela A, Sridharan S, Sciascia N, et al. · Molecular cell 2016What this evidence cannot show
A method paper. The sensitivity figure is for engineered, synchronous nuclease cuts; a stochastic endogenous break spread over a kilobase is a harder target than a restriction site.
Evidence that argues against this (1)
- The same paper is explicit that chromatin features, not sequence preference alone, dictate the repertoire of breaks these enzymes make, and that it detects breaks 'distributed over extended regions' during class switch recombination. A base-pair-resolution instrument does not make a diffuse lesion into a point lesion, so a hotspot called from END-seq inherits the width of the underlying biology rather than the resolution of the assay. Canela A et al. 2016
loop-anchor fragility is the structural model the BCR architecture entry asks to be tested against; not yet measured at this locus
cohesin binding, not transcription, determines where TOP2 is trapped; transcription accelerates conversion to an irreversible break
converts the transcription-factory idea into a mappable CTCF/cohesin loop-anchor feature
- related to →CTCF/cohesin boundaries and the HOXA topological domain — the 3D scaffold the fusion program runs on
same CTCF/cohesin machinery, opposite role: boundary function downstream of the fusion versus breakage upstream of it
- refutes →Apoptotic nucleases cut KMT2A, and cells that survive the attempt can carry a translocation
a TOP2/loop-anchor account cannot be complete: apoptotic cleavage of the KMT2A BCR is merbarone-insensitive and episome-sufficient
TDP2 activity is required for conversion of trapped TOP2cc into an irreversible loop-anchor DSB
asks whether the KMT2A BCR overlaps a called CTCF/cohesin loop anchor in a haematopoietic progenitor
the hypothesis is the locus-level application of loop-anchor fragility
re-analysis of deposited break maps over the KMT2A interval is the cheapest test of the loop-anchor claim at this locus
CTCF/cohesin loop anchors are vulnerable to TOP2B-mediated DSBs
CTCF/cohesin-bound loop anchors are the fragile sites that mechanism describes
a loader degron scored for BREAKAGE rather than transcription is the unperformed experiment
makes the within-side transcription-dependence contradiction decide the question: the G1 explanation for the HCT116 non-hotspot works under Gothe and fails under Canela 2017
opposite direction for CTCF at KMT2A break sites: there CTCF/cohesin anchors ARE the break hotspots, here CTCF occupancy prevents rearrangement. Recorded as an unresolved dispute, not a correction
the rival reading of the same address: a CTCF/cohesin loop anchor where TOP2 breaks concentrate, rather than a protective insulator
the two Canela papers this entry holds are absent from all 81 of Gray et al.'s references
adjudicates side A: the loop-anchor fragility model, as proposition (i)
places the KMT2A element in the CTCF+RAD21 co-occupied class that entry's model is about
names the deposited files behind the 2019 half of the loop-anchor fragility literature
at this element the deposited DRB arm sides with Gothe: TK6 falls from the 98.0th to the 71.0th chr11 percentile on elongation inhibition
2019 is the paper with the directed cohesin experiments; 2017 has the cell-type and species breadth and a Mus spretus occupancy-polymorphism arm
enumerates the deposited files and the full-text cell panel behind the Gothe half of the loop-anchor dispute
assay timings in this dispute span 30 min to two days after release, a 96-fold range treated as one readout
reconciles the within-side contradiction: occupancy sets WHERE, transcription sets HOW MUCH, so Canela and Gothe describe different quantities