Oncofusion

start / graph / mechanism-kmt2a-aff1-dot1l-recruitment-route

Mechanismstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q-partner-provenance

KMT2A::AFF1 has no mapped interface to DOT1L: it lacks both structurally defined recruitment modules, and the one complex-level route claimed for it is contested by reconstitution

also called how does MLL-AF4 reach DOT1L · index fusion DOT1L interface gap · AHD-free H3K79 methylation

There are two known protein hooks by which a leukaemia fusion can grab the enzyme that writes the H3K79 tag, and the infant fusion MLL-AF4 has neither. Some work says its elongation complex binds the enzyme anyway, other work says the opposite, and nobody has pinned the contact to a specific piece of the fusion.

The same thing, for a scientist

**The checkable statement.** Two protein interfaces are documented to recruit DOT1L to a KMT2A fusion: (i) the **ANC1 homology domain (AHD)** of AF9/ENL, an intrinsically disordered module that binds DOT1L directly [E1][E2]; and (ii) the **AF10 octapeptide-motif / leucine-zipper region**, which UniProt annotates on P55197 as residues 703-784, *'Transactivation domain; required for DOT1L-binding'*, with the leucine zipper at 750-778 [E3]. **KMT2A::AFF1 carries neither.** AFF1 (P51825) cross-references exactly three Pfam families - PF05110 (AF-4), PF18875 (AF4_int), PF18876 (AFF4_CHD) - and neither PF17793 (AHD, present on both MLLT3 and MLLT1) nor any leucine-zipper family [E4]. **The one contact it does have points the wrong way.** AFF1 retains PF18875/IPR043639, the AF4 interaction motif (this graph places the conserved core at aa 761-775 and the NMR construct 2LM0 at 738-779), and that motif's job is to bind the AF9/ENL **AHD** - so the index fusion can reach an AHD *in trans*. But the AHD is a single competitive surface: it binds AF4, its paralogue AFF4, **or** DOT1L [E5], and in reconstitution *'DOT1 forms a stable complex with AF9 and does not interact with AF9-AF4-P-TEFb complexes'* [E6]. Occupying the AHD with AFF1 is therefore a route to P-TEFb, not a route to DOT1L. **There is a claimed route, and it is complex-level rather than mapped.** The AFF1/AFF4 super elongation complex is described as binding histone methyltransferases including DOT1L, NSD1 and CARM1 [E11], which would give the index fusion a route without any AFF1 domain contacting DOT1L directly. That claim and Biswas' reconstitution negative [E6] are not reconcilable as stated, and neither has been tested in a t(4;11) cell; what is missing in both readings is a residue-level contact on AFF1. Separately, the AF4-AF9 interface is functionally required in t(4;11) cells: a peptide disrupting it kills MLL-AF4 leukaemia cells while sparing normal haematopoietic progenitors [E12] - so the interface the index fusion does have is load-bearing, whatever it delivers. **And the phenotype is index-established.** H3K79 profiles distinguish human MLL-rearranged ALL and DOT1L suppression silences MLL-AF4 target genes [E7]; MLL-AF4 spreading co-occurs with H3K79me2/3 and confers DOT1L-inhibitor sensitivity [E8]. So this is not a claim that KMT2A::AFF1 lacks H3K79 methylation. It is a claim that **the graph asserts a recruitment mechanism the index fusion has no stated route to execute**, and that the mechanism it asserts was measured in fusions that do have the interface.

Go deeper
The interface ledger
fusionDOT1L-recruiting element carried by the 3' partnerevidence
KMT2A::MLLT3 (MLL-AF9)AHD (PF17793), binds DOT1L at three sites; graded mutation gives graded H3K79me2/3 loss12
KMT2A::MLLT1 (MLL-ENL)AHD (PF17793)45
KMT2A::MLLT10 / PICALM::MLLT10OM-LZ, P55197 703-784 'required for DOT1L-binding'3
KMT2A::AFF1 (MLL-AF4)none - AF4_int (PF18875) binds the AHD, it is not an AHD456

The asymmetry is not subtle and it is not a matter of degree: the two partners with a DOT1L interface have one, and the index partner does not.

What is left unexplained, stated as rival models

Something puts H3K79me2/3 on KMT2A::AFF1 target genes. Four models remain, and this graph currently asserts none of them explicitly:

  1. In-trans AHD with a spare surface. The fusion binds one AF9/ENL molecule through AF4_int while a different AF9/ENL molecule, or a different DOT1L-binding site on the same protein, carries DOT1L. DOT1L has three AF9-binding sites 2, so stoichiometry is not obviously prohibitive - but Biswas' reconstitution says the AF4-occupied complex does not contain DOT1 6, which is the direct test and it is negative.
  2. No targeting at all - DOT1L follows the polymerase. If the fusion raises elongation, H3K79me2 accumulates across the gene body because that is where DOT1L already works, and the mark is a consequence of transcription rather than of recruitment. This predicts H3K79me2 gain should track nascent transcription gain quantitatively and should not require any fusion-DOT1L contact.
  3. The endogenous, AF10-directed unit acting on genes the fusion opened. In the CRISPR-ChIP partition this is the unit menin inhibition is predicted to spare 9. If it is the unit operating at KMT2A::AFF1 targets, then menin inhibition in t(4;11) disease should barely move H3K79me2, and the 'oncogenic DOT1L unit' has no counterpart in the index fusion.
  4. An unmapped AFF1 contact. AFF1 is ~87% disordered, which is exactly the architecture in which a short linear motif hides from domain annotation. Pfam absence is weak evidence for a disordered protein.

Models 1-4 make different predictions and a single degron-plus-calibrated-ChIP experiment separates them (exp-ahd-occupancy-competition-in-t411-cells).

The claimed route, stated fairly

This entry's claim is that no interface is mapped, not that no route has been proposed. Two proposals exist and both are complex-level.

  1. The SEC carries DOT1L. AF4/AFF1 and AF5/AFF4 are described as the backbone of super elongation complexes whose functions include binding histone methyltransferases 'like DOT1L, NSD1 and CARM1', generating H3K79me2/3 and H3K36me2 signatures 11. If correct, KMT2A::AFF1 reaches DOT1L simply by being an SEC component, with no AFF1 domain touching DOT1L. What is missing: which AFF1 residues, at what stoichiometry, and how this survives the reconstitution result that AF9-AF4-P-TEFb complexes contain no DOT1 6. The two reports are in direct tension and the reconciliation, if there is one, is probably that 'SEC' names a family of compositionally distinct assemblies rather than one complex.

  2. AEP and DOT1L cooperate at the same genes. Demonstrated - but in MLL-ENL and MLL-AF10, both of which carry their own DOT1L interface 10.

Neither proposal is refuted here. The point is that a claim this graph has been treating as mechanism rests on a complex-level assertion in tension with the only reconstitution that tested it, and that settling it requires a t(4;11) experiment nobody has run.

A paper whose title is more index-friendly than its experiments

Okuda et al. 2017 is titled 'Cooperative gene activation by AF4 and DOT1L drives MLL-rearranged leukemia' and concludes with a rationale for 'simultaneous inhibition of the MLL fusion-AF4 complex and DOT1L'. It is the closest thing in the literature to a mechanistic account of AF4 and DOT1L acting together, and it is not yet cited anywhere in this graph 10.

But the fusions it uses are MLL-ENL and MLL-AF10 - both of which carry a DOT1L interface. 'AF4' in that title is AEP, the complex, recruited by ENL-family proteins; it is not the MLL-AF4 fusion. An entry that cited this paper as index-fusion evidence for AF4-DOT1L cooperation would be making exactly the import this beat exists to count, and would be doing it from a title rather than from a design.

Why this matters to the menin-interface bet

The menin-interface hypothesis is not damaged by this gap - menin binds the KMT2A N-terminus, which every fusion retains, so it is the one part of the story that is partner-independent by construction. The damage is to the chain that links menin inhibition to H3K79 loss to cell death. That chain has a hole in the index fusion at the step where DOT1L is recruited, and the CRISPR-ChIP result which was supposed to close it turns out to be a mouse MLL-AF9 measurement (ana-crispr-chip-fusion-identity-resolved).

Stated as a falsifier: if DOT1L inhibition and menin inhibition are non-synergistic in human KMT2A::AFF1 cells while being synergistic in MLL-AF9 cells, the two-unit model is AF9-specific and the graph's account of why menin inhibitors work in infant disease needs a different middle.

Correction to this entry, 2026-09-19

This entry was first written in this cycle asserting that KMT2A::AFF1 has no documented route to DOT1L. Two sources found later the same day make that too strong: an AFF1-interactome study describing the SEC as binding DOT1L 11, and a t(4;11) functional study showing the AF4-AF9 interface is required for leukaemia-cell survival 12. The title, both summaries and the section structure were revised from no documented route to no mapped interface, and the claimed route is now stated in its own section rather than only in counter-evidence. The interface ledger and the rival-models analysis are unchanged, because neither source maps a contact on AFF1.

Facts and figures
uniprotP51825
unexplainedthe residue-level contact between KMT2A::AFF1 and DOT1L; complex-level routes are claimed and contested
interfaces present on index fusion
[
  "AF4_int / PF18875 / IPR043639 (binds the AF9-ENL AHD)",
  "AFF4_CHD / PF18876 (AFF-family dimerisation)",
  "AF-4 / PF05110 (N-terminal, largely lost at the breakpoint)"
]
dot1l recruiting interfaces absent from index fusion
[
  "AHD / PF17793 (AF9, ENL)",
  "OM-LZ / P55197 703-784 + 750-778 (AF10)"
]
rival models
[
  "in-trans AHD with a spare DOT1L site",
  "no targeting - DOT1L follows elongating Pol II",
  "the endogenous MLLT10-directed unit acting on fusion-opened genes",
  "an unmapped short linear motif in disordered AFF1"
]
claimed but unmapped routes
[
  "AFF1/AFF4 super elongation complex described as binding DOT1L at complex level (PMID 26171280, weak, no DOI, in tension with PMID 21896721)",
  "AEP and DOT1L cooperation demonstrated in MLL-ENL and MLL-AF10 (PMID 28394257), both of which carry their own DOT1L interface"
]
How this could bite us back
  • If DOT1L reaches KMT2A::AFF1 targets by following the polymerase rather than by fusion recruitment, then a DOT1L inhibitor in t(4;11) disease is acting on a downstream consequence and combination logic built on 'blocking recruitment' is misconceived - which would help explain why pinometostat's single-agent activity was modest.
  • Conversely, if the AF4_int-AHD contact is on the causal path, any inhibitor of that protein-protein interaction removes P-TEFb and DOT1L together, and P-TEFb is required by essentially every transcribed gene, so the therapeutic window is set by general elongation toxicity rather than by fusion selectivity.
  • The absence of a mapped interface is also an absence of a drug target: this graph cannot nominate a KMT2A::AFF1-DOT1L interaction inhibitor because there is no interaction to inhibit.
Still open
  1. Can the tension between an SEC that binds DOT1L (PMID 26171280) and a reconstitution in which AF9-AF4-P-TEFb contains no DOT1 (PMID 21896721) be resolved by compositional heterogeneity of the SEC, and has anyone purified AFF1-containing complexes from a t(4;11) cell and asked whether DOT1L is present?
  2. Does PFWT-mediated disruption of AF4-AF9 change H3K79 methylation in t(4;11) cells? The 2004 study read out apoptosis; the chromatin readout would discriminate delivery from displacement.
  3. Does DOT1L co-immunoprecipitate with KMT2A::AFF1 from a human t(4;11) line under conditions where it co-immunoprecipitates with MLL-AF9 from an AML line? A side-by-side in the same hands has not, to this beat's reading, been published.
  4. Can an AF4_int point mutant that abolishes AHD binding be introduced into KMT2A::AFF1 without abolishing P-TEFb recruitment? If not, the AHD route cannot be tested cleanly in cells and the question must be answered in vitro.
  5. Is H3K79me2 gain at KMT2A::AFF1 targets quantitatively proportional to nascent transcription gain? That single measurement discriminates the recruitment model from the transcription-coupled model.
Every source behind this entry (12)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The AHD of AF9 is intrinsically disordered and recruits multiple partners, including AF4 and DOT1L, through coupled folding and binding, with conserved aliphatic residues of each partner forming part of the complex's hydrophobic core.

    PMID 23260655DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. NMR on isolated domains and peptides; the structure actually solved here is AF9 AHD with AF4, and the DOT1L complex comes from the same group's follow-up. The AHD is a partner's domain, not the fusion's: MLL-AF9 has it because AF9 is the partner.

    Evidence that argues against this (1)
    • Disorder means the AHD can adopt different conformations for different partners, so 'same hydrophobic core' does not strictly prove one binding site; a model in which AF4 and DOT1L occupy overlapping but non-identical surfaces would permit ternary complexes at low occupancy. Kuntimaddi A et al. 2015
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    DOT1L has three AF9-binding sites; an NMR structure of a DOT1L-AF9 complex plus structure-guided point mutations gives graded reduction of DOT1L recruitment to MLL-AF9, graded loss of H3K79me2 and me3 at MLL-AF9 target genes, and a corresponding graded loss of transformation.

    Degree of recruitment of DOT1L to MLL-AF9 defines level of H3K79 Di- and tri-methylation on target genes and transformation potential. · Kuntimaddi A, Achille NJ, Thorpe J, Lokken AA, Singh R, Hemenway CS, Adli M, Zeleznik-Le NJ, Bushweller JH · Cell Rep 2015
    PMID 25921540DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. Entirely MLL-AF9: this is the paper that establishes the AHD-DOT1L axis quantitatively, and it establishes it in the partner KMT2A::AFF1 does not have. The dose-response between recruitment and transformation is exactly the relationship that has never been measured for the index fusion, because there is no interface to titrate.

    Evidence that argues against this (1)
    • If H3K79 methylation scales with DOT1L recruitment to the fusion, then a fusion with no DOT1L interface should carry little ectopic H3K79me2 - yet human MLL-AF4 leukaemias have distinguishing H3K79 profiles. Either the graded relationship is AF9-specific, or recruitment is not the only route to the mark. Krivtsov AV et al. 2008
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProtKB P55197 (primary gene name MLLT10, 1,068 aa) annotates residues 703-784 as 'Transactivation domain; required for DOT1L-binding' and residues 750-778 as a leucine zipper, and annotates no such region on AFF1.

    What this evidence cannot show

    Accession fetched directly on 2026-09-19 and the primary gene name and length read back rather than resolved by a gene_exact query. UniProt regions are curatorial calls drawn from primary literature; the boundaries 703-784 and 750-778 are annotation, not a crystallographic footprint, and the supporting experiments are the AF10/DOT1L coiled-coil literature rather than a single definitive structure.

    Evidence that argues against this (1)
    • The DOT1L-AF10 interaction is not a simple heterodimer: a higher-order configuration of the two coiled-coil domains potentiates leukemogenic activity, so 'AF10 has the interface and AFF1 does not' understates how much of AF10's contribution is oligomeric state rather than binding per se. Song X et al. 2019
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProtKB P51825 (AFF1, 1,210 aa) cross-references only Pfam PF05110 (AF-4), PF18875 (AF4_int) and PF18876 (AFF4_CHD), while P42568 (MLLT3, 568 aa) and Q03111 (MLLT1, 559 aa) each carry PF17793 (AHD) with PF03366 (YEATS); AFF1 carries no AHD and no leucine-zipper family.

    What this evidence cannot show

    Each accession fetched directly on 2026-09-19 and the PRIMARY gene name and length read back. Pfam coverage is model-based: an absent assignment is an absent detected match. UniProt annotates AFF1 as overwhelmingly disordered, and short linear motifs in disordered regions are systematically under-detected by profile HMMs, so this is a weaker negative than the same negative would be for a globular protein.

    Evidence that argues against this (1)
    • AFF1's retained AF4_int motif is itself an example of a short interaction element that Pfam does capture, so the family's coverage of AFF1's binding repertoire is not zero - which makes it harder to argue that a DOT1L-binding motif is simply hiding. UniProt Consortium 2026
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The AHD of AF9/ENL binds AF4, its paralogue AFF4, or the H3K79 methyltransferase DOT1L, and a small molecule that blocks these protein-protein interactions reduces both super-elongation-complex assembly and DOT1L-mediated H3K79 methylation in leukaemia cells.

    Small-molecule inhibitor of AF9/ENL-DOT1L/AF4/AFF4 interactions suppresses malignant gene expression and tumor growth. · Wu F, Nie S, Yao Y, Huo T, Li X, Wu X, Zhao J, Lin YL, Zhang Y, Mo Q, Song Y · Theranostics 2021
    PMID 34373735DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. The 'or' in the abstract's framing is the authors' description of a shared site; the paper's primary contribution is a micromolar chemical probe (IC50 0.9-3.5 uM), not a structural demonstration of exclusivity. A single compound that reduces both SEC and H3K79me cannot distinguish blocking two independent interactions from blocking one shared one.

    Evidence that argues against this (1)
    • A compound at micromolar potency that suppresses onco-MLL, DOT1L, HOXA9 and MYC signatures simultaneously is exactly the profile a non-selective agent produces; ascribing the H3K79 effect specifically to AHD occupancy requires a separation-of-function control this abstract does not describe. Kuntimaddi A et al. 2015
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In reconstitution from defined components, DOT1 forms a stable complex with AF9 and does not interact with AF9-AF4-P-TEFb complexes; AF4 and AFF4 independently bridge P-TEFb to AF9 through direct contacts with AF9 and cyclin T1.

    Function of leukemogenic mixed lineage leukemia 1 (MLL) fusion proteins through distinct partner protein complexes. · Biswas D, Milne TA, Basrur V, Kim J, Elenitoba-Johnson KS, Allis CD, Roeder RG · Proc Natl Acad Sci U S A 2011
    PMID 21896721DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. Reconstitution with purified minimal complexes 'in the absence of other factors' - a stringent test of sufficiency that can miss an interaction stabilised in cells by chromatin, by a bridging factor, or by the KMT2A moiety itself, none of which are present in the tube. The functional validation in this paper is MLL-AF9 immortalisation, not MLL-AF4.

    Evidence that argues against this (1)
    • The same paper shows these complexes are dynamic and interconverting, so a negative in a reconstituted steady state does not exclude a transient DOT1L visit to an AF4-bound AHD in cells, which is the exact timescale on which a processive methyltransferase would work. Leach BI et al. 2013
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    H3K79 methylation profiles distinguish human MLL-rearranged ALLs from other ALLs and correlate with Mll-AF4-associated expression in a murine model, and DOT1L suppression inhibits expression of critical MLL-AF4 target genes.

    H3K79 methylation profiles define murine and human MLL-AF4 leukemias. · Krivtsov AV, Feng Z, Lemieux ME, Faber J, Vempati S, Sinha AU, Xia X, Jesneck J, Bracken AP, Silverman LB, Kutok JL, Kung AL, Armstrong SA · Cancer Cell 2008
    PMID 18977325DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. ChIP-chip, uncalibrated; 'human MLL-rearranged ALL' is a partner-mixed patient group, so the human half of this claim is not KMT2A::AFF1-specific even though the murine half is. RNAi-era DOT1L suppression.

    Evidence that argues against this (1)
    • The mark being present and functionally required does not establish that the fusion recruits the writer, which is the claim this entry says is unsupported; a transcription-coupled model predicts the same profile without any fusion-DOT1L contact. Biswas D et al. 2011
  8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MLL-AF4 binding spreads into the gene body at a subset of targets, accompanied by menin spreading and increased H3K79me2/3, and the expression of those spreading genes is preferentially downregulated by DOT1L inhibition relative to other H3K79me2/3-marked genes.

    MLL-AF4 Spreading Identifies Binding Sites that Are Distinct from Super-Enhancers and that Govern Sensitivity to DOT1L Inhibition in Leukemia. · Kerry J, Godfrey L, Repapi E, Tapia M, Blackledge NP, Ma H, Ballabio E, O'Byrne S, Ponthan F, Heidenreich O, Roy A, Roberts I, Konopleva M, Klose RJ, Geng H, Milne TA · Cell Rep 2017
    PMID 28076791DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. Human MLL-AF4 ChIP-seq; the abstract does not name the cell lines, so the breadth of the t(4;11) evidence is not bounded by this citation. Occupancy correlation, not a recruitment assay. **Resolved favourably:** the same hedge can be dropped. GSE83671 contains **30 of 59 samples SEM (t(4;11), the index fusion)**, plus ML-2, THP-1, KOPN-8, MV4-11, CCRF-CEM, RCH-ACV, a presentation-ALL PDX, cord blood, second-trimester fetal bone marrow and an EPZ-5676 dose series. Resolved 2026-09-19 by the Q18b unstated-systems sweep from the study's deposition, not from its text; bounded as a statement about the deposited data rather than about the whole study.

    Evidence that argues against this (2)
    • That menin spreading and H3K79me2/3 track MLL-AF4 spreading together is equally consistent with both following elevated transcription, and the DOT1L-inhibitor contrast is a sensitivity difference among marked genes rather than evidence about how the mark arrived. Godfrey L et al. 2019
    • Thirty of fifty-nine samples being one cell line means this is index-fusion evidence concentrated in SEM rather than distributed across KMT2A::AFF1 systems. It is much better than the hedge implied and still not a survey. NCBI Gene Expression Omnibus 2026
  9. E9established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    CRISPR-ChIP partitions H3K79 methylation into an oncogenic DOT1L unit directed by the MLL fusion in a menin-dependent manner and a separate endogenous unit directed by MLLT10.

    CRISPR-ChIP reveals selective regulation of H3K79me2 by Menin in MLL leukemia. · Gilan O, Talarmain L, Bell CC, Neville D, Knezevic K, Ferguson DT, Boudes M, Chan YC, Davidovich C, Lam EYN, Dawson MA · Nat Struct Mol Biol 2023
    PMID 37679565DOI
    What this evidence cannot show

    Abstract read from the efetch record on 2026-09-19; full text unreachable (Europe PMC not OA, Unpaywall closed, bioRxiv HTTP 429 on four attempts). Confidence is reduced from 'strong' to 'moderate' for THIS entry's purpose because the deposited data show every H3K79 experiment is murine MLL-AF9, and the relevance here is to KMT2A::AFF1 - see ana-crispr-chip-fusion-identity-resolved. **Regraded moderate -> established on 2026-09-20** under `concept-the-confidence-grading-rule`: the paper's own reported partitioning, graded established elsewhere.

    Evidence that argues against this (1)
    • If the oncogenic unit is defined by direction from a fusion that carries an AHD, there may be no oncogenic unit at all in a fusion that carries none - in which case KMT2A::AFF1 H3K79 methylation is entirely the 'endogenous' MLLT10-directed activity, and menin inhibition should not move it. Deshpande AJ et al. 2014
  10. E10strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MLL-ENL and MLL-AF10 constitutively activate transcription by inducing both AEP-dependent activation and DOT1L-dependent maintenance, mostly in the absence of PRC1, to fully transform haematopoietic progenitors, which the authors present as a rationale for simultaneous inhibition of the MLL fusion-AF4 complex and DOT1L.

    Cooperative gene activation by AF4 and DOT1L drives MLL-rearranged leukemia. · Okuda H, Stanojevic B, Kanai A, Kawamura T, Takahashi S, Matsui H, Takaori-Kondo A, Yokoyama A · J Clin Invest 2017
    PMID 28394257DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. Cited here for what it does NOT show: despite naming AF4 in the title, the fusions tested are MLL-ENL and MLL-AF10, both of which carry a DOT1L interface, and 'AF4' denotes the AEP complex recruited by ENL-family proteins rather than the MLL-AF4 fusion. Murine haematopoietic progenitor transformation. This paper is cited nowhere else in this graph as of cycle 49.

    Evidence that argues against this (1)
    • The cooperation demonstrated here is the strongest published argument that AEP and DOT1L act on the same genes, which is the premise the index fusion would need - so reading it as irrelevant to KMT2A::AFF1 may be too strict. What it cannot do is establish the contact, because in both fusions tested the contact is the partner's own AHD or OM-LZ. Yokoyama A et al. 2010
  11. E11weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    AF4/AFF1 and AF5/AFF4 form the backbone of super elongation complexes whose two main functions are recruiting P-TEFb from inhibitory 7SK RNPs and binding histone methyltransferases including DOT1L, NSD1 and CARM1, giving transcribed genes H3K79me2/3 and H3K36me2 signatures; the study maps the protein-protein interaction network of the AF4 and AF4N complexes and reports AF4N additionally recruiting TFIIH and MEN1.

    AF4 and AF4N protein complexes: recruitment of P-TEFb kinase, their interactome and potential functions. · Scholz B, Kowarz E, Rössler T, Ahmad K, Steinhilber D, Marschalek R · Am J Blood Res 2015
    PMID 26171280
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read. PubMed's esummary returns NO DOI for this record, and the journal (Am J Blood Res) is low-visibility; the interaction network is from overexpression and co-purification rather than from reconstitution, and the DOT1L statement appears in the framing rather than as a mapped result in the abstract. Confidence is set to weak for exactly these reasons. This paper is cited nowhere else in this graph.

    Evidence that argues against this (1)
    • Reconstitution from defined components finds that DOT1 binds AF9 but not AF9-AF4-P-TEFb complexes, which is the direct opposite of an SEC that carries DOT1L. One possible reconciliation is that 'SEC' denotes compositionally distinct assemblies and that the DOT1L-containing and AF4-containing ones are different; on that reading both reports can stand and neither describes a contact on AFF1. Biswas D et al. 2011
  12. E12moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    A synthetic peptide (PFWT) designed from mapping of the AF9-binding domain of AF4 disrupts the AF4-AF9 interaction in vitro and in vivo, inhibits proliferation of t(4;11) leukaemia cells expressing MLL-AF4 through apoptosis, and does not affect the proliferative capacity of haematopoietic progenitor cells.

    The synthetic peptide PFWT disrupts AF4-AF9 protein complexes and induces apoptosis in t(4;11) leukemia cells. · Srinivasan RS, Nesbit JB, Marrero L, Erfurth F, LaRussa VF, Hemenway CS · Leukemia 2004
    PMID 15269783DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, every identifier copied from that record. A 2004 peptide study: cell-permeable peptides of that era commonly act through membrane or off-target effects, the abstract gives no concentration, and 'does not affect' normal progenitors is stated without a quantitative window. Crucially it establishes that the AF4-AF9 interface matters in t(4;11) cells, NOT that the interface delivers DOT1L - the readout is apoptosis, not H3K79 methylation.

    Evidence that argues against this (1)
    • If disrupting AF4-AF9 kills t(4;11) cells, and if the same surface is the one DOT1L uses, then a compound blocking it removes the fusion's P-TEFb route and any putative DOT1L route at once - so the result cannot distinguish which of the two the cells were dying for, and it is equally consistent with the entry's claim that the interface delivers P-TEFb rather than DOT1L. Wu F et al. 2021
Connected to