CALM-AF10 needs BMI1 - and so does MLL-AF9, so this is a shared polycomb dependency and not a point of difference (title corrected 2026-09-21)
also called CALM-AF10 BMI1 dependency
CALM-AF10 leukaemia cells depend on a protein called BMI1 that keeps stem-cell genes switched off and stem-cell behaviour switched on. Removing it, or blocking it with a drug, stops the leukaemia in mice.
The same thing, for a scientist
**BMI1 is consistently overexpressed in adult and paediatric CALM-AF10-positive leukaemias, and genetic Bmi1 depletion abrogates CALM-AF10-mediated transformation of murine HSPCs.** CALM-AF10-positive murine and human AML cells are sensitive to the BMI1 inhibitor PTC-209 and to PTC-596, and PTC-596 significantly prolongs survival of mice xenografted with a human CALM-AF10 line [E1]. The BMI1 overexpression was visible in patient T-ALL profiling two decades ago, alongside HOXA [E2], and the 2026 review names high BMI1 as a point where PICALM::MLLT10 **differs** from KMT2A-rearranged leukaemia, which shares the HOXA and MEIS elevation but not the BMI1 elevation [E3]. **CORRECTION, 2026-09-21 (Q67):** the fusion-distinguishing half of that claim does not survive. Bmi1 is indispensable for the development of MLL-AF9 leukaemia in vivo - Bmi1-null granulocyte/macrophage progenitors transformed with MLL-AF9 are still immortalised in vitro but differentiate more, retain fewer leukaemia stem cells and fail to make leukaemia in transplant, with Ink4a/Arf deletion only partially restoring it [see E3 counter-evidence]. So BMI1 requirement is SHARED between a KMT2A fusion and CALM-AF10 in mouse models. What survives, and the distinction is the whole point: **EXPRESSION differs and REQUIREMENT does not.** The patient-profiling study [E2] reports BMI1 over-expressed in CALM-AF10-positive T-ALL and NOT in MLL-translocated acute leukaemia, and BMI1 lies close to the AF10 breakpoint - so a difference in level is measured and has a candidate cis explanation. A difference in level is not a difference in dependency, and the dependency experiment has now been run on both sides with the same answer. BMI1 also remains a second dependency orthogonal to DOT1L and to menin, with chemical matter in clinical development. Note that the level claim is itself cohort-dependent: in an unselected paediatric T-ALL series, BMI1 was high across all phenotypic subgroups with no CALM-AF10 cases present at all.
How much weight the 'differs from KMT2A-r' claim can carry - none, as of 2026-09-21
The differentiating claim came from a review 3 citing a patient-profiling study 2, not from a head-to-head comparison of BMI1 dependency across fusions. It has now been contradicted rather than merely left unmeasured. Bmi1 is indispensable for MLL-AF9 leukaemia development in vivo, and the Bmi1-null phenotype in that system - enhanced differentiation, fewer retained leukaemia stem cells, derepression of candidate tumour suppressors, only partial rescue by Ink4a/Arf deletion - is the same shape as the Bmi1 requirement reported for CALM-AF10 transformation.
So the honest position is now: BMI1 EXPRESSION is reported higher in CALM-AF10-positive T-ALL than in MLL-translocated acute leukaemia in the profiling study, and BMI1 sits close to the AF10 breakpoint, which is a plausible cis explanation. BMI1 REQUIREMENT is established in murine CALM-AF10 transformation AND in murine MLL-AF9 leukaemia. So the level differs and the dependency does not, and the entry's original title collapsed the two. The level claim is also cohort-dependent: in an unselected paediatric T-ALL series BMI1 was high across all phenotypic subgroups, with no CALM-AF10 cases in the cohort at all.
The experiment still worth doing is the quantitative one: BMI1 protein level and PTC-596 dose-response in matched CALM-AF10 and KMT2A-rearranged models, which would test whether a measured difference in level translates into a therapeutic window. The qualitative one - is BMI1 required in both - has an answer, and it is yes.
| mechanism class | polycomb dependency |
|---|---|
| gene | BMI1 (PCGF4), canonical PRC1 |
| evidence type | expression in patients; genetic requirement in murine transformation; pharmacological sensitivity in murine and human cells |
| shared with kmt2a rearranged | claimed not shared, by a review rather than by a head-to-head experiment |
| orthogonality | independent of both the DOT1L axis and the menin axis; a CALM-AF10 cell could in principle be attacked through BMI1 whether or not it is menin-dependent |
chemical matter
[ "PTC-209", "PTC-596" ]
- BMI1 is required for normal haematopoietic stem-cell self-renewal, so a BMI1-directed agent carries an on-target stem-cell toxicity that is not obviously separable from the therapeutic effect.
- PTC-596's mechanism is contested; a clinical programme built on 'BMI1 inhibitor' framing may be running a tubulin agent under a targeted label.
- Does the measured difference in BMI1 EXPRESSION between CALM-AF10 and MLL-translocated leukaemia translate into a difference in sensitivity to BMI1 inhibition? The qualitative dependency question is now answered - both require Bmi1 in mice - so what remains is a dose-response comparison, and it has not been done.
- Is the PTC-596 xenograft survival benefit a BMI1 effect or a tubulin effect? Without a BMI1-null control arm the compound result cannot distinguish them.
- Does DOT1L inhibition combine with BMI1 inhibition? Both are reported active in this fusion separately, and the combination has not been tested.
- If BMI1 is required on both sides of this graph's comparison, what is left that IS fusion-distinguishing? Every mechanism this beat has examined - DOT1L recruitment, HOXA output, polycomb dependency - now has a KMT2A-rearranged counterpart.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
BMI1 is consistently overexpressed in adult and paediatric CALM-AF10-positive leukaemias; genetic Bmi1 depletion abrogates CALM-AF10-mediated transformation of murine HSPCs; CALM-AF10-positive murine and human AML cells are sensitive to PTC-209 and PTC-596; and PTC-596 prolongs survival of mice injected with a human CALM-AF10 cell line.
Acute myeloid leukemia driven by the CALM-AF10 fusion gene is dependent on BMI1. · Barbosa K, et al. · Exp Hematol 2019What this evidence cannot show
Murine transformation plus a human CALM-AF10 xenograft. PTC-209 and PTC-596 are not BMI1-selective by mechanism - PTC-596 is a microtubule-targeting agent whose BMI1 effect is downstream - so the pharmacological arm is weaker than the genetic one. The human line used is a myeloid CALM-AF10 line, while most patients have T-ALL.
Evidence that argues against this (1)
- The genetic and pharmacological arms do not measure the same thing: PTC-596's antiproliferative activity is now widely attributed to tubulin binding rather than to BMI1 degradation, so xenograft survival under PTC-596 is not a clean BMI1 result. Barbosa K et al. 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
CALM-AF10-positive T-ALL expression profiles are characterised by overexpression of HOXA and BMI1 oncogenes.
CALM-AF10+ T-ALL expression profiles are characterized by overexpression of HOXA and BMI1 oncogenes. · Dik WA, et al. · Leukemia 2005What this evidence cannot show
Patient expression profiling; associative. Small numbers of CALM-AF10 cases.
Evidence that argues against this (1)
- PUBLISHED CONFLICTING RESULT that bounds how far BMI1 elevation discriminates. This study's contrast is CALM-AF10-positive T-ALL against other T-ALL and against MLL-translocated acute leukaemia, and it is real. But in a separate unselected paediatric T-ALL series of 43 patients, upregulated BMI1 was observed by quantitative real-time PCR in ALL PHENOTYPIC SUBGROUPS - and none of those patients carried the CALM-AF10 fusion transcript. So high BMI1 is not rare in fusion-negative T-ALL, and how much of the CALM-AF10 contrast is fusion-driven rather than cohort- or platform-driven has not been settled. The comparison that would settle it - BMI1 level in CALM-AF10-positive versus CALM-AF10-negative T-ALL, matched for maturation stage and measured on one platform - is in neither paper. Sayitoğlu M et al. 2012
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review
PICALM::MLLT10 leukaemias share HOXA and MEIS elevation with KMT2A-rearranged leukaemias but additionally show high BMI1 expression, and BMI1 is required for CALM-AF10 transformation.
PICALM::MLLT10 translocated leukemia. · Cullen JM, et al. · FEBS Lett 2026What this evidence cannot show
REVIEW. The 'additionally' is the review's framing of two separate primary results rather than a measured head-to-head comparison; no matched BMI1-dependency experiment across fusions exists in this beat's reading.
Evidence that argues against this (1)
- DIRECT PUBLISHED CONTRADICTION OF THIS ENTRY'S TITLE. BMI1 dependence is NOT a feature that KMT2A-rearranged leukaemia lacks: Bmi1-null granulocyte/macrophage progenitors transformed with MLL-AF9 were still immortalised in vitro but showed enhanced differentiation, retained fewer leukaemia stem cells, and Bmi1 was INDISPENSABLE for the development of leukaemia in vivo - with deletion of both Ink4a and Arf only partially restoring leukaemogenic capacity, and re-complementation with Bmi1 failing to restore most deregulated genes or in vivo leukaemogenicity. A KMT2A fusion and CALM-AF10 both require Bmi1 for leukaemia in a mouse. The polycomb dependency is shared, and this entry's framing of it as a point of difference between the arms does not survive. Yuan J et al. 2011
an orthogonal dependency of this fusion
Bmi1 depletion abrogates CALM-AF10 transformation
CALM-AF10 needs BMI1 and, on transplantation, so does MLL-AF9 - so BMI1 is a shared polycomb dependency of two HOXA-driven lesions rather than a point of difference or a KMT2A-specific bypass
removes a claimed discriminator: the polycomb dependency the comparison arm was said to have uniquely is present on both sides