The menin arm and the CTCF arm are not two stories: menin/SETD2/H3K36me3/LEDGF licenses CtIP short-range resection and CTCF/H3K9me/HP1-gamma licenses BRCA1/BARD1-EXO1/DNA2 long-range resection, which are the two consecutive halves of ONE reaction - so a menin inhibitor and an H3K9-axis inhibitor are acting on the same pathway at different steps

{
  "both_cascades_real_as_published": "~0.8 each - both are peer-reviewed in Nucleic Acids Research with mapped interfaces and separation-of-function reagents (BARD1 PEELI; menin Val371-Gln400)",
  "both_operate_in_a_haematopoietic_cell": "LOW and entirely unmeasured. The CTCF paper is U2OS and HEK-293T; the menin paper's in vivo arm is a Men1-ablation LUNG carcinogenesis model. This is where most of the prior is lost.",
  "consecutive_rather_than_redundant_parallel": "~0.4. The CTCF paper's own discussion describes the recruitment routes as parallel, and its epistasis rests on NON-ADDITIVITY of co-depletions, which a shared ceiling or incomplete knockdown also produces.",
  "product": "~0.15, dominated by the cell-type term"
}

[
  "In a KMT2A-rearranged human B-lineage line, menin inhibition and CTCF or HP1-gamma depletion are SUB-ADDITIVE on DR-GFP HR efficiency: the double perturbation lands within 1.2-fold of the deeper single arm.",
  "Menin inhibition impairs SHORT-range resection (ssDNA at ~0.3-1.6 kb from an AsiSI site) while sparing the incremental long-range signal, and CTCF or HP1-gamma depletion does the reverse - impairing the increment between ~1.6 kb and ~3.5 kb while leaving short-range initiation nearer normal.",
  "H3K36me3 at a break is NOT lost when CTCF is depleted, and H3K9me3 at a break is NOT lost when menin is depleted - i.e. the two chromatin licences are independently established even though their effectors are sequential.",
  "Because CtIP is required for the short-range step that precedes the long-range one, CtIP depletion is epistatic to BOTH arms: it should abolish long-range resection as completely as it abolishes short-range.",
  "A menin inhibitor's effect on HR, if any, is pocket-independent: an MBM-cavity inhibitor at full target engagement produces a smaller HR deficit than MEN1 knockdown, because the gammaH2A.X-binding surface is Val371-Gln400 and not the cavity.",
  "Prophylactic G9a/GLP inhibition and a menin inhibitor given together during etoposide produce a larger repair deficit than either alone would predict from independence - the clinically relevant corollary and the one this graph should care about most."
]

[
  "exp-hr-competence-under-menin-inhibition-in-kmt2a-rearranged-blasts"
]

[
  "ana-q32-depmap-across-the-hr-licensing-cascade",
  "absence-no-resection-factor-localised-at-the-kmt2a-locus"
]

[
  "FALSIFIER 1 - ADDITIVITY KILLS 'CONSECUTIVE'. System: SEM or RS4;11 (human KMT2A::AFF1 B-ALL), DR-GFP reporter, I-SceI induction, at least three independent transfections. Perturbations: revumenib at a concentration giving documented target engagement; siCTCF or siCBX3; and the combination. SCORING: if the double perturbation reduces HR efficiency to a value at least 1.5-fold BELOW the deeper single arm, with non-overlapping 95% confidence intervals, the two arms are acting additively and are therefore not consecutive steps of one reaction. Sub-additivity is defined in advance as the double landing within 1.2-fold of the deeper single. Anything between 1.2- and 1.5-fold is recorded as unresolved, not as support.",
  "FALSIFIER 2 - THE MENIN ARM DOES NOT OPERATE IN A BLOOD CELL. System: SEM, RS4;11 or MV4-11, DR-GFP reporter, I-SceI, at least three independent experiments. Perturbation: revumenib at a clinically achievable exposure, plus a MEN1 knockdown ceiling control. SCORING: if MEN1 KNOCKDOWN (not merely the drug) leaves HR efficiency within 10% of vehicle, with a 95% confidence interval excluding a 20% reduction, then menin does not license resection in the cells this graph is about, and the menin half of this hypothesis is dead regardless of what happens in lung. This also settles `exp-hr-competence-under-menin-inhibition-in-kmt2a-rearranged-blasts`, which is currently a proposal.",
  "FALSIFIER 3 - NEITHER ARM REACHES THIS LOCUS. Interval: the KMT2A exon-11 CTCF/cohesin element, GRCh38 chr11:118,488,159-118,489,206, inside KMT2A at GRCh38 chr11:118,436,456-118,526,832 (verified against Ensembl ENSG00000118058 on 2026-09-22). Cell type: human CD34+ haematopoietic progenitors or a B-lineage progenitor line. Design: induce a double-strand break within 2 kb of the element (Cas9 or an integrated AsiSI/FokI site), ChIP for BRCA1, CBX3 and CtIP at 5, 15, 30 and 60 min. SCORING: if NONE of the three reaches a fold-enrichment of 1.5 over a matched chr11 control region 20 kb distal, at any timepoint, in at least three biological replicates, then no licensing cascade operates at this element in a blood cell and the entire construction is irrelevant to this disease however well established it is in U2OS.",
  "FALSIFIER 4 - THE TWO ARMS ARE REDUNDANT, NOT SEQUENTIAL. System: AsiSI/DIvA U2OS (to stay in the system where both cascades can be assayed), ssDNA resection qPCR at 335 bp, 1,618 bp and 3,500 bp from a cleaved AsiSI site, at least three independent inductions. SCORING: if menin depletion and CTCF depletion each reduce resection at the SAME distance by a comparable amount - both reducing the 335 bp and 1,618 bp signals by 60% or more while neither selectively spares the 1,618-to-3,500 bp increment - then both arms act at the short-range step and they are redundant rather than consecutive. The hypothesis specifically predicts a DISSOCIATION between the two distances by arm, and its absence refutes it.",
  "FALSIFIER 5 - ONE ARM IS DOWNSTREAM OF THE OTHER, SO THERE ARE NOT TWO ARMS. System: U2OS with laser micro-irradiation or a FokI site, at least three independent experiments. SCORING: if CTCF depletion abolishes H3K36me3 accumulation at the break (reduction of 50% or more in strip intensity relative to siCTL), or menin depletion abolishes H3K9me3 accumulation at the break by the same criterion, then the two chromatin licences are not independently established and the 'two arms, two steps' framing collapses into a single linear cascade - which would be a different and simpler claim than this hypothesis makes."
]

{
  "for_combination_therapy": "A menin inhibitor and an H3K9-axis inhibitor would be synthetic on DNA repair. This graph proposes each separately, for unrelated reasons, in patients who also receive etoposide, daunorubicin or doxorubicin. Nobody has considered them together and the combination is not obviously irrational as a leukaemia therapy - which is exactly why the repair liability needs to be on record before somebody proposes it.",
  "for_the_menin_interface_hypothesis": "It gives menin a second, damage-conditional, NON-TRANSCRIPTIONAL function with its own mapped surface - Val371-Gln400 for gammaH2A.X - which is not the MBM cavity that revumenib, ziftomenib and bleximenib occupy. That is a separation-of-function opportunity and a warning at once: a pocket inhibitor might spare it, while a menin degrader would not.",
  "for_the_genesis_argument": "It would make the accuracy of repair at a TOP2 break at the KMT2A locus depend on two chromatin states, only one of which (H3K9/CTCF) this graph has ever measured there. H3K36me3 at the KMT2A breakpoint cluster region has never been measured in any cell type as far as this beat found, and it would become a first-order question.",
  "for_the_existing_Q110_hypothesis": "Nothing is withdrawn. `hyp-the-writer-reaches-the-kmt2a-anchor-only-through-ctcf-and-only-after-the-break` stands and is strengthened by the mechanism entry written this beat. This hypothesis sits beside it and concerns a different pair of arms."
}