Oncofusion

start / graph / gene-nipbl

Geneweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q43-break-repair-machinery

NIPBL - the cohesin loader, the cleanest degron result in the whole architecture literature, and a second bounded absence in this disease

also called NIPBL · IDN3 · SCC2 · delangin · Nipped-B-like protein

NIPBL is the machine that loads the cohesin ring onto DNA, so without it the loops do not form at all. Deleting it in mouse liver made every chromosome loop vanish and changed almost no gene's expression - which is the single most sobering result anyone building a therapy on loop architecture has to answer.

The same thing, for a scientist

NIPBL (Q6KC79, 2804 aa) is the cohesin loader, required for cohesin's association with chromatin and therefore upstream of every loop anchor in this beat. It earns an entry for one reason and it is not a disease reason: **Nipbl deletion in mouse liver makes TADs and Hi-C peaks vanish globally in the absence of transcriptional changes** [E2]. That result, with the parallel cohesin-degron and CTCF-degron results, is the baseline against which every architectural claim in this graph has to be read, and `genome-ctcf-cohesin-hoxa-tad-boundary` already carries it as a standing caution. **In this disease there is nothing.** `(MLL OR KMT2A) AND NIPBL` returns 13 records and title-checking finds **0** about KMT2A-rearranged leukaemia - ten are Cornelia de Lange syndrome or developmental-genetics papers in which NIPBL is the principal cohesinopathy gene and 'MLL' appears via KMT2A/KMT2D on the same diagnostic panel [E3]. The complex form was run as Rule 64 requires and it surfaces cohesin-level KMT2A literature that no subunit or loader symbol reaches [E4], but none of it perturbs or measures NIPBL. So this entry is a **bounded absence plus a load-bearing negative from another system**, and that combination is more useful than either alone: the loop machinery is central to the break model, and the one experiment that removed loops entirely changed nothing transcriptionally.

Go deeper
What the Nipbl result does and does not bound

It bounds transcription. Removing loops did not change gene expression in mouse liver, so a claim that a leukaemic transcriptional programme depends on a TAD boundary has to explain why this locus is an exception to a global negative. genome-ctcf-cohesin-hoxa-tad-boundary states this as a standing caution against the HOTTIP/CTCF boundary results and this entry agrees.

It does not bound breakage. Nothing in the degron literature measured DNA breaks. Cohesin binding is a stated precondition for TOP2 trapping 5, and loop anchors accumulate TOP2B-mediated DSBs, so a NIPBL or cohesin degron could in principle abolish loop-anchor fragility while leaving transcription untouched. That is a specific, cheap and unperformed experiment, and it is the sharpest thing this entry contributes: the architecture field's flagship negative is about the wrong readout for this beat.

The absence, written so it can be re-run
querycounton-beat after title check
(MLL OR KMT2A) AND NIPBL130
(MLL OR KMT2A) AND cohesin174
KMT2A AND "loop extrusion"00

The third row is worth its own line. loop extrusion is the mechanism NIPBL performs and the phrase is the field's standard term; quoted and bounded it returns zero records alongside KMT2A, with a clean querytranslation and no UID artefact 3. Aliases run: NIPBL, IDN3, SCC2, delangin. SCC2 was title-checked rather than counted, because it collides with unrelated usages.

What would change this: any NIPBL or cohesin-loading perturbation in a KMT2A-rearranged system, or any loop-extrusion analysis that includes the KMT2A locus.

Facts and figures
uniprotQ6KC79
length AA2804
gene symbolNIPBL
accession resolutiongene_exact:NIPBL AND organism_id:9606 AND reviewed:true returned exactly 1 row; gene_primary NIPBL, length 2804, synonyms IDN3 SCC2.
functioncohesin loader; with MAU2/SCC4 loads cohesin onto chromatin and is required for loop extrusion
entry typeBOUNDED ABSENCE in this disease, plus a load-bearing negative from another system
unperformed experimentNIPBL or cohesin degron with etoposide, scoring DSBs at loop anchors and specifically at the KMT2A BCR
counted absence
{
  "query": "(MLL OR KMT2A) AND NIPBL",
  "count": 13,
  "on_beat_after_title_check": 0,
  "dominant_topic_of_hits": "Cornelia de Lange syndrome and developmental genetics (10 of 13)",
  "loop_extrusion_query": "KMT2A AND \"loop extrusion\" returns 0, clean querytranslation",
  "date": "2026-09-20",
  "aliases_run": [
    "NIPBL",
    "IDN3",
    "SCC2",
    "delangin"
  ],
  "complex_form_run": "(MLL OR KMT2A) AND cohesin, 17 records, 4 on beat",
  "repo_grep": "NIPBL appears in this graph only in the degron_baseline block of genome-ctcf-cohesin-hoxa-tad-boundary, cited to the mouse-liver result"
}
degron baseline
{
  "nipbl_deletion_mouse_liver": "TADs and Hi-C peaks vanish globally with no transcriptional changes (PMID 29094699)",
  "readout_measured": "transcription",
  "readout_NOT_measured": "DNA double-strand breakage",
  "consequence": "the architecture field's flagship negative uses the wrong readout for this beat; a loader or cohesin degron scored for loop-anchor and KMT2A-BCR breakage has never been run"
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.5897,
  "chronos_mean_other_leukaemia": -0.454,
  "difference": -0.1357,
  "pan_essentiality_all_lines": -0.3251,
  "selectivity_rank": 589,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
How this could bite us back
  • NIPBL is not a target and could not be. Its haploinsufficiency causes Cornelia de Lange syndrome, a multisystem developmental disorder, which is an absolute floor in a paediatric population.
  • This entry's main use is deflationary: the strongest architectural negative in the field says loops can be removed entirely without changing transcription. Any loop-based therapeutic argument in this graph has to answer it, and the honest answer is that nobody knows whether the same is true for breakage because nobody measured breakage.
  • An absence of KMT2A-specific NIPBL literature is an absence of testing, not a negative result. Nothing here should be cited as evidence that the loader is irrelevant to this disease.
Still open
  1. Does removing cohesin loading abolish loop-anchor DNA breakage? The flagship degron experiments in this field all measured transcription and none measured breaks, and that single omission is why the architecture-versus-fragility question is still open.
  2. Is the mouse-liver transcriptional null generalisable to a proliferating fetal haematopoietic progenitor? Liver is post-mitotic with a static programme, which is the tissue least likely to show a transcriptional requirement for loops.
  3. Does NIPBL occupancy differ at the KMT2A BCR between fetal and adult progenitors? Cohesin loading sets where cohesin sits, cohesin binding predicts TOP2 trapping, and nobody has looked at either in the right cell type.
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    NIPBL is UniProt Q6KC79, Nipped-B-like protein (delangin, SCC2 homolog), 2804 residues, primary gene name NIPBL with synonyms IDN3 and SCC2.

    What this evidence cannot show

    A naming fact. gene_exact:NIPBL AND organism_id:9606 AND reviewed:true returned exactly one row with gene_primary NIPBL. At 2804 aa NIPBL is the longest protein in this beat and has no close paralogue, so the length check is unusually informative here - unlike the SMC1A/SMC3 pair 16 residues apart.

    Evidence that argues against this (1)
    • NIPBL functions with MAU2 (SCC4) as an obligate heterodimer, so a NIPBL perturbation is a loader-complex perturbation. MAU2 was not resolved in this session and nothing here covers it; the same naming-versus-measurement caution that applies to the cohesin subunits applies to the loader. UniProt Consortium 2026
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Deletion of Nipbl in mouse liver causes topologically associating domains and Hi-C peaks to vanish globally in the absence of transcriptional changes.

    Two independent modes of chromatin organization revealed by cohesin removal. · Schwarzer W, Abdennur N, Goloborodko A, Pekowska A, Fudenberg G, Loe-Mie Y, et al. · Nature 2017
    PMID 29094699DOI
    What this evidence cannot show

    Mouse liver, a post-mitotic differentiated tissue with a largely static transcriptional programme - which is the most important caveat on generalising the transcriptional null. A tissue in which most genes are stably expressed is the tissue least likely to reveal a transcriptional requirement for loops. It also measures transcription, not DNA breakage.

    Evidence that argues against this (1)
    • The locus-specific results in this disease run the other way: HOTTIP loss alters the HOXA-driven TAD and its transcription, boundary CBS deletion reduces promoter/enhancer interactions and target expression and mitigates leukaemogenesis in xenografts. A global transcriptional null and a locus-specific transcriptional requirement are not formally contradictory, but this graph's standing position is that the locus-specific claims must be read against the global baseline rather than in isolation. Luo H et al. 2022
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    As of 2026-09-20 the PubMed query (MLL OR KMT2A) AND NIPBL returns 13 records, none of which title-checks as being about KMT2A-rearranged leukaemia - ten are Cornelia de Lange syndrome or developmental-genetics studies - and the query KMT2A AND "loop extrusion" returns 0 records with a clean querytranslation containing no UID parsing artefact.

    What this evidence cannot show

    A counted absence. Aliases run: NIPBL, IDN3, SCC2, delangin. SCC2 was title-checked rather than counted because of collisions. The repository was grepped for NIPBL before this was written: it appears only in the degron_baseline block of genome-ctcf-cohesin-hoxa-tad-boundary, cited to the mouse liver result and not to anything in this disease.

    Evidence that argues against this (1)
    • The complex form was run as Rule 64 requires and it does change the picture at the complex level - '(MLL OR KMT2A) AND cohesin' surfaces a KMT2A-specific study of cohesin loss that no subunit or loader symbol reaches. The NIPBL-level absence survives that check; it would not have survived it unchecked, and this project has twice published a confident wrong absence by skipping it. Meyer A et al. 2024
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.preprint

    Cohesin loss and MLL-AF9 are not synthetic lethal in murine haematopoietic stem and progenitor cells: complete Stag2 loss and Smc3 haploinsufficiency were well tolerated in MLL-AF9-expressing cells with minimal effect on in-vitro self-renewal.

    PMID 38352423DOI
    What this evidence cannot show

    A preprint read from its abstract; murine; in-vitro self-renewal endpoint; the perturbations are Stag2 and Smc3, not NIPBL. It is cited here as the nearest KMT2A-specific cohesin evidence and as the demonstration that the complex form reaches literature the loader symbol does not.

    Evidence that argues against this (1)
    • Reducing a cohesin subunit is not the same as removing the loader. NIPBL deletion abolishes cohesin loading entirely and produces a global architectural phenotype, whereas Smc3 haploinsufficiency leaves a functional ring at lower dose. A tolerated haploinsufficiency says nothing about loader ablation. Schwarzer W et al. 2017
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    TOP2 chromatin localisation and etoposide trapping require pre-existing binding of cohesin to DNA, and cohesin binding with transcriptional activity alone predicts 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, Wutz G, Tang W, Zagnoli-Vieira G, et al. · Mol Cell 2019
    PMID 31202577DOI
    What this evidence cannot show

    Genome-wide in cultured cells, complex-level, and the KMT2A BCR is not examined specifically. It is cited here to establish that removing cohesin loading is a plausible way to abolish loop-anchor fragility - the experiment this entry argues has never been done.

    Evidence that argues against this (1)
    • Even if a loader degron abolished loop-anchor breakage, it would not follow that this explains KMT2A breakage in a fetus, where there is no etoposide and where the graph's break-machinery beat finds breaks occurring in untreated cytokine-stimulated CD34+ cells with no genotoxin at all. Canela A et al. 2017
Connected to