Oncofusion

start / graph / exp-menin-inhibitor-dose-response-calm-af10-panel

Experimentmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q55-calm-af10

Run the menin inhibitor against CALM-AF10 next to a KMT2A-rearranged control - the missing two-week experiment

also called the CALM-AF10 menin experiment

Take leukaemia cells carrying the CALM-AF10 fusion, treat them with a menin-blocking drug alongside cells carrying a KMT2A fusion, and see whether the HOXA genes switch off. Nobody has published this, and it would take about two weeks.

The same thing, for a scientist

A matched-exposure comparison that settles `hyp-menin-axis-separable-from-hoxa-programme`. **Index arm:** U-937 (human, CALM-AF10, myeloid) plus at least one CALM-AF10 T-lineage model, because 65% of patients are T-ALL and the only available human line is myeloid. **Positive control arm, run in the same plate on the same day:** a KMT2A-rearranged line (MV4-11 or MOLM-13). **Readouts in order of value:** (1) HOXA9/MEIS1 transcript by RT-qPCR at 48-96 h; (2) viability/proliferation over 10-14 days, since menin-inhibitor cytotoxicity in KMT2A-rearranged cells is slow and a 72-hour assay would give a false negative *in both arms*; (3) H3K79me2 by calibrated ChIP at HOXA, which distinguishes a menin-independent MLLT10-directed unit from a menin-dependent fusion-directed one. **Target-engagement control is the non-negotiable part**: without demonstrating menin occupancy in the CALM-AF10 arm, a negative is uninterpretable.

Go deeper
Why this is worth doing before anything else in this beat

It is the only experiment that converts a well-bounded absence into a fact, it needs no new reagents, and either answer is publishable and changes this graph. If CALM-AF10 is insensitive, menin dependence is a property of the KMT2A-menin contact rather than of the HOXA programme, and this project's central hypothesis becomes sharper. If it is sensitive, menin dependence does not require a menin-binding motif, and a large part of the graph's mechanistic story needs rewriting - which is worth knowing at any price.

The controls that decide whether a negative means anything
  1. Target engagement in the index arm. Menin occupancy or a menin-inhibitor pharmacodynamic marker must be shown in the CALM-AF10 cells themselves. A drug that does not enter the cell produces the same result as a drug the cell does not need.
  2. Same plate, same day, same lot. The KMT2A-rearranged control is the calibration, not a historical number.
  3. Long enough. Menin-inhibitor killing of KMT2A-rearranged cells is a differentiation phenotype over 10-14 days; a 3-day CellTiter-Glo is the standard way to produce a false negative in both arms and then report a 'specificity' difference that is really an assay-duration artefact.
  4. A DOT1L-inhibitor arm. CALM-AF10 cells SHOULD respond to a DOT1L inhibitor. If they do not, the model is wrong and the menin negative is uninterpretable. This is the positive control for the biology rather than for the drug.
  5. Genotype the line. U-937's CALM-AF10 status should be confirmed by RT-PCR or RNA-seq in the actual vial used, not assumed from the catalogue.
The harder second half

A negative in the index arm leaves open the indirect route: CALM-AF10 might still need the wild-type menin-KMT2A1 complex to establish HOXA even if it does not need it to maintain it. Two additions settle it: MEN1 CRISPR knockout (genetic, no exposure question) and KMT2A knockout in the same line. If both are tolerated while DOT1L knockout is not, the separation is complete.

Facts and figures
designMatched-exposure dose-response of a menin inhibitor (revumenib or VTP-50469) in CALM-AF10 cells (U-937 plus a T-lineage CALM-AF10 model) against a KMT2A-rearranged line (MV4-11 or MOLM-13) run in the same experiment, with a DOT1L-inhibitor arm as the biological positive control and a menin target-engagement readout in the index arm.
readoutHOXA9 and MEIS1 transcript at 48-96 h; viability and differentiation over 10-14 days; calibrated H3K79me2 ChIP at HOXA; optional MEN1 and KMT2A CRISPR knockout arms.
controlsTarget engagement in the index arm; same-plate KMT2A-rearranged comparator; 10-14 day duration; DOT1L-inhibitor positive control; fusion genotype confirmed in the vial used.
cost tierlow
timeline2-3 weeks for the pharmacological arms; 6-8 weeks including the CRISPR knockout arms.
what a negative meansOnly interpretable with target engagement shown. With it, a negative separates menin dependence from the HOXA programme.
what a positive meansMenin dependence does not require a menin-binding motif, and the indirect wild-type-KMT2A1 route is operating in CALM-AF10 as it does in NPM1c and NUP98-rearranged disease.
How this could bite us back
  • A three-day viability assay would give a negative in BOTH arms and be reported as a specificity result. The duration is the experiment.
  • If the index arm is only U-937, the answer is about a myeloid line in a T-lineage-skewed disease, and the claim must say so.
Still open
  1. Which T-lineage CALM-AF10 model? The disease is T-ALL-skewed and the standard human line is myeloid; a patient-derived xenograft may be the only honest T-lineage instrument.
  2. What exposure counts as full menin target engagement in a cell line that does not depend on menin? The usual pharmacodynamic markers (MEIS1, MEF2C down) are themselves the endpoint here, which is circular - an occupancy assay is needed instead.
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    No published study exposes a CALM-AF10 system to a menin inhibitor or removes MEN1 from one, as of 2026-09-20.

    What this evidence cannot show

    Bounded absence with operand controls and two same-shape positive controls; see ana-calm-af10-menin-bounded-absence.

    Evidence that argues against this (1)
    • If a CALM-AF10 patient has been treated on AUGMENT-101 cohort 2D, an unreported clinical instance of this experiment already exists. ClinicalTrials.gov 2026
  2. E2weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset

    In DepMap, the one screened CALM-AF10 line has MEN1 gene effect -0.229 while KMT2A-rearranged lines average -1.263, giving a concrete prior for the expected result.

    What this evidence cannot show

    n=1, myeloid line only, no statistic computed. This sizes the experiment rather than substituting for it.

    Evidence that argues against this (1)
    • DepMap barely resolves HOXA9 dependency in any leukaemia arm (KMT2A-rearranged mean -0.271 against an other-leukaemia mean of -0.221), so a CRISPR screen is a poor instrument for the HOXA-programme half of the question; the transcript readout in this experiment is doing work the screen cannot. DepMap / Open Targets Platform 2026
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    CALM-AF10-transformed cells respond to the DOT1L inhibitor EPZ004777 with loss of Hoxa and Meis1 expression and selectively impaired proliferation, which is the biological positive control this experiment requires.

    PMID 23138183DOI
    What this evidence cannot show

    Murine. In a human line the equivalent arm should use pinometostat or a current DOT1L inhibitor at a documented exposure.

    Evidence that argues against this (1)
    • DOT1L inhibition worked in murine CALM-AF10 models and failed clinically in KMT2A-rearranged disease - and the pinometostat trials enrolled on '11q23 rearrangement', which excludes t(10;11)(p13;q14) PICALM::MLLT10 because that breakpoint is at 11q14. So the preclinical promise and the clinical failure were never in the same disease. ClinicalTrials.gov 2026
Connected to