Experiment — is a KMT2A-septin fusion menin-dependent? The one-plate version
also called septin fusion menin dependence assay
A small, cheap experiment: grow mouse blood progenitors carrying a KMT2A-septin fusion, add a menin-blocking drug, and see whether the colonies stop forming and the HOXA9 signal drops. Nobody has done it.
The same thing, for a scientist
Serial replating colony assay of mouse bone marrow progenitors transduced with MLL-SEPT6 versus MLL-AF9, treated with a menin inhibitor (revumenib or VTP-50469) across a dose range, with a genetic Men1 deletion arm as the orthogonal control and pinometostat as the partner-mechanism comparator. Readouts: colony number at rounds 2-3, Hoxa9 and Meis1 by qPCR, menin and fusion occupancy at Hoxa9 by ChIP, and H3K79me2 over the Hoxa cluster. The design exists because the constructs exist: the MLL-SEPT6 retroviral construct series including Δcoil, ΔGTP, ΔP-loop, S56N and Δcoil-ER is published, and an MLL/SEPT6-immortalised IL-3-dependent mouse line (HF6) is described in the literature [E1][E4]. The discriminating comparison is not the absolute menin-inhibitor effect but the RATIO of menin-inhibitor to DOT1L-inhibitor sensitivity between the septin and elongation arms.
Design
Arms. (1) MLL-SEPT6; (2) MLL-AF9 (elongation-partner positive control, the standard menin-inhibitor-sensitive background); (3) MLL-SEPT6 Δcoil (non-transforming negative control, confirms the assay reads fusion-driven self-renewal 1); (4) empty vector.
Treatments. Menin inhibitor dose range; pinometostat dose range; vehicle. Orthogonal genetic arm: Men1 floxed progenitors with tamoxifen-inducible Cre, which removes the possibility that an effect is off-target pharmacology 2.
Readouts. Colony number and morphology at rounds 2 and 3; Hoxa7, Hoxa9 and Meis1 by qPCR (all three are documented as up in MLL-SEPT6 cells, so the assay has a defined positive signal 1); menin ChIP at Hoxa9; H3K79me2 ChIP across the Hoxa cluster.
The comparison that matters. The oligomerisation-first hypothesis predicts the menin-inhibitor dose-response curves for arms 1 and 2 are similar while the pinometostat curves diverge, with the septin arm relatively resistant. The standard elongation-hijack model predicts both drugs behave similarly across both arms. A septin arm resistant to BOTH drugs would refute the menin thesis for this partner class and is the result with the largest clinical consequence.
Why this has not been done, and why that is not a good reason
There is no human KMT2A-septin cell line in the standard panels or in DepMap, so the experiment has to be built in mouse progenitors rather than bought. That is a week of work with published constructs, not a programme. The literature contains zero records intersecting menin with septin terms against denominators of 1264 and 1936 3, so the barrier is attention rather than difficulty.
| design | Serial replating colony assay of mouse bone marrow progenitors transduced with MLL-SEPT6, MLL-AF9, MLL-SEPT6 Δcoil or empty vector, treated with a menin inhibitor and, in parallel, with pinometostat, across dose ranges; orthogonal genetic arm using Men1-floxed progenitors with inducible Cre. |
|---|---|
| readout | Colony number and morphology at replating rounds 2-3; Hoxa7/Hoxa9/Meis1 qPCR; menin ChIP at Hoxa9; H3K79me2 ChIP across the Hoxa cluster. Primary comparison: the RATIO of menin-inhibitor to DOT1L-inhibitor sensitivity between septin and elongation arms. |
| controls | MLL-AF9 as the menin-inhibitor-sensitive elongation-partner positive control; MLL-SEPT6 Δcoil as the non-transforming negative control; empty vector; genetic Men1 ablation as the orthogonal control for off-target pharmacology. |
| cost tier | low |
| timeline | 6-10 weeks including construct rebuild; 3-4 weeks if the published constructs are obtained. |
| decision rule | If the menin-inhibitor dose-response of the septin arm matches the elongation arm while the pinometostat response diverges, the oligomerisation-first reframing is supported and menin's partner-independence is demonstrated rather than inferred. If the septin arm is menin-resistant, the menin thesis has a documented partner-class boundary and this project's target strategy needs revision. |
tests hypotheses
[ "hyp-elongation-hijack-is-a-special-case-of-forced-oligomerisation" ]
- A negative result in colonies would be ambiguous because MLL-SEPT6 needs a cooperating lesion for acute disease; the experiment is only decisive in the positive direction unless the in vivo arm is run.
- If the septin arm turns out menin-resistant, an approved drug is being given to a patient class on a mechanism that does not apply, and this experiment would be the thing that showed it. That is a reason to run it early, not a reason to avoid it.
- Should the in vivo arm (MLL-SEPT6 plus FLT3-ITD, menin inhibitor in transplanted mice) be run in parallel? The colony assay measures immortalisation and the disease model needs a cooperating lesion, so a colony-only negative would be ambiguous.
- Is there a human KMT2A-septin patient sample anywhere — a biobank, a case-report centre — that could be put on a short-term ex vivo menin-inhibitor assay? That would be worth more than the whole mouse design.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The MLL-SEPT6 retroviral construct series (wild type, Δcoil, Δcoil-ER, ΔGTP, ΔP-loop, S56N) and a serial myeloid immortalisation assay in mouse bone marrow progenitors are published, with intact MLL-SEPT6 immortalising and every mutant failing, and Hoxa7, Hoxa9 and Meis1 upregulated in the immortalised cells.
Dimerization of MLL fusion proteins and FLT3 activation synergize to induce multiple-lineage leukemogenesis. · Ono R, Nakajima H, Ozaki K, et al. · J Clin Invest 2005What this evidence cannot show
Constructs described in a 2005 paper; current availability from the originating laboratory is not established and they may need to be rebuilt, which is straightforward from the published design.
Evidence that argues against this (1)
- MLL-SEPT6 alone produced only long-latency myeloproliferative disease in vivo, so a colony assay measures immortalisation rather than leukaemia; a negative menin-inhibitor result in colonies would not exclude menin dependence of the established disease, and the in vivo arm (FLT3-ITD co-expression) would be needed to settle it. Ono R et al. 2005
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Menin is essential for maintenance of MLL-associated but not other oncogene-induced myeloid transformation, and acute genetic ablation of menin reverses aberrant Hox gene expression and abrogates the differentiation arrest and oncogenic properties of MLL-transformed blasts — providing the genetic control arm this design uses.
The menin tumor suppressor protein is an essential oncogenic cofactor for MLL-associated leukemogenesis. · Yokoyama A, Somervaille TC, Smith KS, et al. · Cell 2005What this evidence cannot show
The abstract does not specify which fusions were used; this entry cites the paper for the existence and behaviour of the conditional-ablation approach, not for any partner-specific claim.
Evidence that argues against this (1)
- A conditional-knockout control does not resolve degree: KMT2A-rearranged ALL and AML respond to revumenib with different kinetics and different phenotypic outcomes in vitro, so the same genetic arm can look different depending on lineage, and a septin arm's behaviour would need lineage-matched comparators. Adriaanse FRS et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
PubMed [Title/Abstract] searches return 1264 records for menin and 1936 for septin OR septins but exactly 0 for their intersection, and 0 for (MLL OR KMT2A) AND septin AND (menin OR MEN1 OR DOT1L OR H3K79 OR super elongation OR SEC).
NCBI E-utilities esearch over PubMed, [Title/Abstract]-scoped Boolean queries run for this beat on 2026-09-19 with both halves of each intersection counted separately · US National Library of Medicine, PubMed · PubMed (E-utilities) 2026What this evidence cannot show
Title/abstract scoping; an experiment reported only in a supplementary figure would not be found, and negative results of exactly this kind are systematically unpublished.
Evidence that argues against this (1)
- The pivotal revumenib KMT2A-r trial had an unknown fusion partner in 22.8% of its efficacy population, so septin-partner patients may already have been treated with a menin inhibitor without anyone being able to say so; the literature absence is about named topics, not about clinical exposure. Issa GC et al. 2025
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
HF6, an IL-3-dependent mouse bone-marrow-derived cell line immortalised by MLL/SEPT6 and fully transformable by class I mutations such as FLT3 mutants, has been used as a screening background.
Possible involvement of RasGRP4 in leukemogenesis. · Watanabe-Okochi N, Oki T, Komeno Y, et al. · Int J Hematol 2009What this evidence cannot show
Described incidentally in a screening paper; provenance and availability not established here.
Evidence that argues against this (1)
- An IL-3-dependent line is a poor setting for a menin-inhibitor dose-response because growth-factor signalling can sustain proliferation independently of the fusion's transcriptional output, which would produce a false negative; primary-progenitor colony assays are the more interpretable readout. Ono R et al. 2005
the discriminating experiment
the fusion to test, because it is the only one with published constructs
menin-inhibitor arm
- tests →Pinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
DOT1L comparator arm
the experiment this entry proposes rests on an unreadable founding paper