The two founding conditional Mll knockouts were published back to back and they do not agree: fatal marrow failure in three weeks, or grossly normal blood
also called Jude versus McMahon · the Mll conditional knockout discrepancy
Two labs removed the same gene from the blood system of mice and published side by side in one journal issue: one saw blood production collapse within three weeks, the other saw blood that looked normal. Both are still cited, usually without mentioning the other.
The same thing, for a scientist
Cell Stem Cell 1(3), September 2007, carries two independent conditional Mll alleles. Jude et al. report that Mll is essential for maintenance of adult HSCs and progenitors, with FATAL BONE MARROW FAILURE within 3 weeks of deletion [E1]. McMahon et al. report that mice with Mll conditionally deleted in the haematopoietic system have GROSSLY NORMAL haematopoiesis in bone marrow, thymus and spleen, with the defect appearing only on competitive transplantation [E2]. Both agree that Mll-deficient cells fail in competition and both agree that the quiescent HSC fraction falls. The disagreement is confined to whether steady-state haematopoiesis survives deletion - and that is precisely the parameter a therapeutic-index argument needs. This graph's KMT2A entry has never named either paper, so it has also never named the discrepancy.
What each paper says, in its own words
Jude et al. 1: 'Mll is essential for the maintenance of adult HSCs and progenitors, with fatal bone marrow failure occurring within 3 weeks of Mll deletion.'
McMahon et al. 2: 'Mice with Mll expression conditionally deleted in the hematopoietic system have grossly normal hematopoiesis in bone marrow, thymus, and spleen. However, transplanted Mll-deficient bone marrow cells are highly compromised in their ability to competitively reconstitute irradiated recipients.'
These are not compatible readings of the same phenotype at the level of the animal.
What they agree on, which is the part worth building on
- Competitive failure. Both report that Mll-deficient cells lose in competition - Jude in mixed bone marrow chimeras, McMahon in competitive reconstitution of irradiated recipients 12.
- Loss of quiescence. Jude: ectopic cell-cycle entry and depletion of quiescent HSCs in the adult. McMahon: the quiescent fraction of fetal HSCs greatly reduced 12. Two alleles, two developmental stages, same direction.
So the replicated results are self-renewal under competition and loss of quiescence. The unreplicated result is steady-state marrow failure.
Why the difference probably is not a contradiction, and why that does not make it safe to ignore
Candidate reconciliations, none of which this entry can settle from abstracts: the two alleles delete different exons and may leave different residual products; the Cre drivers and induction schedules differ; the animals were followed for different periods; and 'grossly normal' is a descriptive call on counts and histology, which a competitive assay is far more sensitive than.
The reason it matters anyway is that the menin-inhibitor therapeutic-index argument in gene-men1 has exactly this shape - steady-state blood is preserved, stress response is not - and it cites the menin knockout (PMID 19228930) for it. If the Mll literature shows the same phenotype can be read as 'fatal marrow failure' by a second lab with a second allele, then the reassuring reading of the menin data inherits that uncertainty. The argument is not refuted; its confidence interval is wider than the citation count suggests.
The failure this entry actually records
Before this cycle neither PMID 18371366 nor PMID 18371367 appeared anywhere in this corpus - not as a citation and not in prose - measured over the 2,396 node files outside this beat's own folder on 2026-09-23 (analyses/q127-the-pi-dimensions-at-the-centre/results.json). The two founding conditional knockouts of the gene this project is named after had zero mentions here.
That is not a literature gap; it is an entry gap, and it is the subject of concept-a-pi-dimension-can-be-saturated-in-the-corpus-and-absent-from-the-entry-named-for-it.
| corpus status before this cycle | PMID 18371366 and PMID 18371367: zero mentions of either, as a citation or in prose, across the 2,396 node files outside this beat's folder, measured 2026-09-23. Persisted in analyses/q127-the-pi-dimensions-at-the-centre/results.json. |
|---|---|
| why it bears on the menin argument | gene-men1's 'why blocking menin does not wreck normal blood' section rests on the same phenotype shape - preserved steady state, impaired stress response - drawn from a single menin knockout (PMID 19228930). The Mll pair shows that this shape is allele-dependent and readable two ways, so the menin reading should carry the same caution. |
disagreement
{
"parameter": "steady-state adult haematopoiesis after pan-haematopoietic Mll deletion",
"jude_2007_pmid_18371366": "fatal bone marrow failure within 3 weeks",
"mcmahon_2007_pmid_18371367": "grossly normal in bone marrow, thymus and spleen",
"agreed": [
"failure to self-renew under competition",
"loss of the quiescent HSC fraction"
],
"same_journal_issue": "Cell Stem Cell volume 1 issue 3, 13 September 2007, pages 324-337 and 338-345",
"resolution_status": "UNRESOLVED HERE. Neither full text was opened; allele design, Cre driver, induction schedule and follow-up duration are the four candidate explanations and none has been checked."
}- If the severe reading is the correct one, then any agent that removes wild-type KMT2A function from the marrow - including a degrader, and conceivably a menin inhibitor at sufficient occupancy - has a three-week marrow-failure liability that no current preclinical package would detect at a steady-state blood count.
- Which exons does each allele delete, and does either leave a translated fragment? Two full texts and an afternoon would resolve most of this and the answer changes how both papers should be cited.
- Has anyone re-derived a conditional Mll allele since 2007 and reported the steady-state phenotype? If a third allele exists it breaks the tie, and nobody in this corpus has looked.
- Does the menin conditional knockout show a steady-state phenotype on a longer follow-up than was published? The reassuring half of the menin therapeutic-index argument rests on a negative at one timepoint.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Using an inducible knockout model, Mll is essential for the maintenance of adult HSCs and progenitors, with fatal bone marrow failure occurring within 3 weeks of Mll deletion, and Mll-deficient cells are selectively lost from mixed bone marrow chimeras.
Unique and independent roles for MLL in adult hematopoietic stem cells and progenitors. · Jude CD, Climer L, Xu D, Artinger E, Fisher JK, Ernst P · Cell Stem Cell 2007What this evidence cannot show
Abstract read, full text not opened; the deleted exons, Cre driver and induction schedule are therefore not recorded here and are the first things a reconciliation would need.
Evidence that argues against this (1)
- Directly contradicted on the steady-state phenotype by the paper published immediately after it in the same issue. McMahon KA et al. 2007
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Mice with Mll expression conditionally deleted in the haematopoietic system have grossly normal haematopoiesis in bone marrow, thymus and spleen, while transplanted Mll-deficient bone marrow cells are highly compromised in competitive reconstitution of irradiated recipients.
Mll has a critical role in fetal and adult hematopoietic stem cell self-renewal. · McMahon KA, Hiew SY, Hadjur S, Veiga-Fernandes H, Menzel U, Price AJ, Kioussis D, Williams O, Brady HJM · Cell Stem Cell 2007What this evidence cannot show
Abstract read, full text not opened. 'Grossly normal' is a qualitative call and the abstract does not say how long the animals were followed, which is the most likely source of the difference from E1.
Evidence that argues against this (1)
- Directly contradicted on the steady-state phenotype by the paper published immediately before it in the same issue. Jude CD et al. 2007
- related to →KMT2A / MLL1 — the gene that breaks
the two founding conditional knockouts of this gene, neither previously cited in the corpus
the replicated half of the pair is the quiescence result
the menin therapeutic-index argument has the same phenotype shape and inherits the same uncertainty