MLLT6 (AF17) — the fourth KMT2A fusion partner inside DotCom, and the only one no cell needs

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

• The DotCom framing suggests that H3K79 methylation is the common route for all KMT2A fusions, but DOT1L inhibition (pinometostat) has already underperformed clinically, so a complex-level account does not by itself produce a target.
• If DotCom and SEC/AEP are alternative assemblies, an agent that displaces one may simply shift the fusion onto the other.
• a partner that moves the drug target - AF17 over-expression drives DOT1L into the cytoplasm by a leptomycin-B-sensitive, XPO1-dependent route. If that holds in a KMT2A::MLLT6 cell, the amount of DOT1L available in the nucleus - and therefore the dose-response of pinometostat-class inhibitors - is partner-dependent. Untested in any leukaemia cell.

1. Does MLLT6 matter in KMT2A::MLLT6 leukemia specifically? No DepMap line carries that fusion, so the only test that would be informative has not been run.
2. Are DotCom and the AF4/ENL higher-order complex mutually exclusive in a living leukemic nucleus, as the in vitro reconstitution implies? The whole question of which assembly a given KMT2A fusion engages turns on this.
3. Why do four of DotCom's subunits, but not the fifth, appear as KMT2A fusion partners? A structural answer would say something about what a partner has to supply.
4. Does AF17 drive DOT1L out of the nucleus in a leukaemia cell, and specifically in a KMT2A::MLLT6 cell? If it does, nuclear DOT1L abundance is fusion-partner-dependent and the failure of pinometostat may be partly a localisation problem rather than only a potency one.

1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

   A multisubunit Dot1 complex (DotCom) was identified containing DOT1L together with ENL, AF9/MLLT3, AF17/MLLT6 and AF10/MLLT10, plus TRRAP, Skp1 and beta-catenin; the human DotCom trimethylates H3K79; and knockdown of Dot1 or of Drosophila homologues of Dot1-associated proteins reduces expression of a subset of Wingless target genes.

   Linking H3K79 trimethylation to Wnt signaling through a novel Dot1-containing complex (DotCom). · Mohan M, Herz HM, Takahashi YH, Lin C, Lai KC, Zhang Y, Washburn MP, Florens L, Shilatifard A · Genes & Development 2010
   PMID 20203130DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. Complex composition from affinity purification plus MudPIT in HeLa cells, with functional work in Drosophila — no leukemia model, no KMT2A fusion protein, and no binary interaction mapping between named subunits. Co-purification does not establish which subunits contact which. INTERACTION PROVENANCE (Q76, 2026-09-21, live IntAct REST), which measures exactly that last sentence: all five IntAct records for the DOT1L-MLLT6 pair are psi-mi:MI:1060 spoke-expanded and typed MI:0914 association, from PMIDs 20203130, 26496610, 28514442 and 33961781, and no PDB entry contains both P55198 and Q8TEK3 - MLLT6 has no PDB entry at all. So AF17's membership of DotCom has never been observed as a binary interaction with DOT1L in any curated record. It remains a real observation that AF17 co-purified with DOT1L in four independent studies, and reproducibility across labs is its own kind of support; what the record cannot support is a direct AF17-DOT1L contact or an interface. Of the four DotCom MLLT proteins, MLLT3 has a direct-interaction record and a structure and MLLT10 has a non-expanded physical-association record and a structure, so MLLT6 is the one whose contact is least constrained. See ana-two-thirds-of-the-curated-interactome-is-a-row-nobody-observed.

   ▸Evidence that argues against this (1)

   • 'A named subunit of DotCom' is a MudPIT co-purification assignment, and co-purification does not distinguish a stoichiometric subunit from a shared or sub-stoichiometric interactor. van Nuland et al. applied label-free quantitative mass spectrometry to the human SET1/MLL complexes precisely to make that distinction, resolving 'both known and novel, unique and shared interactors' and determining their stoichiometry and distribution across complexes; one of their results is that DPY30, long treated as a core COMPASS subunit, is also a genuine subunit of the NURF remodelling complex. The DotCom composition this entry relies on was never quantified that way, so 'named subunit' is a weaker fact than the entry's framing implies - for MLLT6 and for TRRAP, SKP1 and beta-catenin alongside it. Read: abstract in full, not full text. van Nuland R et al. 2013
2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   In reconstitution with defined complexes, DOT1 forms a stable complex with AF9 and does not interact with AF9-AF4-P-TEFb complexes; AF4 and AFF4 independently mediate association of P-TEFb with AF9 through direct interactions with AF9 and cyclin T1; and P-TEFb provides the link for association of ELL, EAF1 and EAF2 with AF4.

   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 KSJ, Allis CD, Roeder RG · Proceedings of the National Academy of Sciences USA 2011
   PMID 21896721DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. Reconstitution with recombinant minimal complexes (insect-cell expression), which establishes what can assemble in vitro rather than what does assemble in a leukemic nucleus. The transformation experiments in this paper use MLL-AF9, not KMT2A::AFF1.

   ▸Evidence that argues against this (1)

   • Reconstitution with defined complexes reports what purified components do at the concentrations supplied, and the negative carried here — that the DOT1-AF9 complex does not interact with AF9-AF4-P-TEFb — is the kind of result most sensitive to that limitation, since a weak or transient bridge is lost on reconstitution. Nature 2012
3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

   MLLT6 has DepMap mean gene effect -0.1054 across 1,208 lines with 1.16% of lines depleted, mean -0.0779 in 11 KMT2A-rearranged leukemia lines against -0.1195 in 43 other leukemia lines (difference +0.042).

   DepMap CRISPR (Chronos) gene effect via the Open Targets Platform GraphQL API, with KMT2A-rearrangement status from Cellosaurus; per-gene recomputation for this beat · Broad Institute DepMap; Cellosaurus; Open Targets Platform; analysis run by this project · Database and derived analysis 2026
   ▸What this evidence cannot show

   DepMap release version not stated by the API. Eleven KMT2A-rearranged lines, none of which is a KMT2A::MLLT6 line, so this does not test whether MLLT6 matters in the fusion it actually participates in.

   ▸Evidence that argues against this (2)

   • The flat DepMap profile (mean -0.105, 1.2% of lines depleted, selectivity +0.042) is presented as 'no dependency signal of any kind', with the subtitle 'the only one no cell needs'. The clinical record disagrees about importance while agreeing about the number. Meyer et al. characterised 760 MLL-rearranged biopsies from 384 paediatric and 376 adult patients at the molecular level and report MLLT6/AF17 among the nine translocation partner genes predominantly involved in MLL recombination. A gene that is a recurrent driver partner in patients and inert in a CRISPR screen is telling you that no screened line carries the relevant fusion - which is exactly what this entry's own open question says - and not that the gene is dispensable to the disease. Read: abstract in full, not full text. Meyer C et al. 2009
   • The germline genetics point the same way and are worth having explicitly, because they change what the DepMap null means. Zhang et al. made the first Af17 mutant mouse: homozygous mutants are viable, fertile, normal in size and show no gross physical, behavioural or haematopoietic abnormality, with only an increase in peripheral-blood neutrophils. Read together with Meyer 2009 this is a therapeutic-window argument, not an irrelevance argument - the profile of a good target in the fusions where MLLT6 is the partner, rather than of a protein that does not matter. Read: abstract in full, not full text. Zhang Z et al. 2010
4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

   In an unbiased genome-wide KMT2A-selectivity scan of 17,728 genes run by this project, MEN1 ranks first with a difference of -0.8656; MLLT1 ranks 487 (difference -0.1492) and MLLT3 ranks 14,902 (difference +0.0521).

   Unbiased genome-wide DepMap scan for KMT2A-rearrangement-selective dependencies, analyses/depmap-kmt2a-selective-scan-full/ in this project (17,728 genes scored, 11 KMT2A-rearranged vs 43 other leukemia lines) · Broad Institute DepMap; Cellosaurus; Open Targets Platform; analysis run by this project · Derived analysis 2026
   ▸What this evidence cannot show

   DepMap release version not stated by the API. Eleven KMT2A-rearranged lines against 43 others; the scan's own caveat about small group sizes applies. MLLT6 itself is not among the scan's reported top-50 or watchlist rows, so its rank is not quoted here.

   ▸Evidence that argues against this (1)

   • A selectivity difference expressed as a rank inherits this project's own correction to its instrument: the scan's statistic divides by comparator standard deviation, which ranks NARROWNESS rather than effect size at Spearman −0.92 in the depleted tail. A rank from that scan is not an effect-size ordering, and 167 entries were written before that was measured. A rank of 487 for MLLT1 in a scaled ranking is also a rank in the part of the distribution where the statistic is least stable. Oncofusion knowledge graph (internal analysis) 2026
5. E_q65_loc1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   In mouse cortical collecting duct M-1 cells, AF17 (MLLT6) over-expression drives preferential CYTOPLASMIC distribution of Dot1a and this is blocked by the nuclear-export inhibitor leptomycin B, while siRNA depletion of AF17 causes nuclear accumulation of Dot1a - making AF17 a measured, CRM1/XPO1-dependent determinant of DOT1L's nucleocytoplasmic partition.

   AF17 facilitates Dot1a nuclear export and upregulates ENaC-mediated Na+ transport in renal collecting duct cells. · Wu H, et al. · PLoS One 2011;6:e27429 2011
   PMID 22087315DOI
   ▸What this evidence cannot show

   Renal epithelium, mouse, over-expression and knockdown rather than endogenous tagging - and NOT leukaemia. Leptomycin B sensitivity identifies the CRM1/XPO1 route but does not prove AF17 is the cargo adaptor. The relevance is by analogy: AF17 is a KMT2A fusion partner in t(11;17)(q23;q21) and DOT1L is this graph's therapeutic target. Abstract only.

   ▸Evidence that argues against this (1)

   • AF17 and AF9 compete for the same Dot1a surface, so an AF17 effect on DOT1L distribution could be an indirect consequence of displacing AF9 rather than of AF17 carrying DOT1L anywhere; the same laboratory's earlier work frames AF17 as a competitor, not a transporter. Mohan M et al. 2010
6. E_q65_loc2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   AF17/MLLT6 is a constitutive subunit of DotCom, a multisubunit DOT1/KMT4 complex that also contains ENL, AF9/MLLT3 and AF10/MLLT10 - three other KMT2A fusion partners - together with TRRAP, Skp1 and beta-catenin, and the human complex trimethylates H3K79.

   Linking H3K79 trimethylation to Wnt signaling through a novel Dot1-containing complex (DotCom). · Mohan M, et al. · Genes Dev 2010;24:574-89 2010
   PMID 20203130DOI
   ▸What this evidence cannot show

   Complex membership is a compositional address, not a subnuclear one, and this entry does not treat it as a localisation. Biochemistry in human cells plus a Drosophila in vivo arm; no leukaemia system. Abstract only.

   ▸Evidence that argues against this (1)

   • If four KMT2A fusion partners are subunits of one DOT1L complex, the partner-specific biology this graph tracks cannot be explained by DotCom membership alone, and a localisation difference between the partners - which nobody has measured - is one of the few remaining places such specificity could live. Wu H et al. 2011
7. E_q69_ct1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   In MICE, Af17 expression is detected in specific cell populations at all embryonic stages examined and in multiple adult tissues by X-gal staining, in situ hybridisation and RT-PCR, and homozygous Af17-mutant mice are viable, fertile, normal in size and display no gross physical, behavioural or haematopoietic abnormality, leading the authors to conclude that Af17 may be dispensable for normal embryogenesis, haematopoiesis and animal survival.

   Widely expressed Af17 is likely not required for embryogenesis, hematopoiesis, and animal survival. · Zhang Z, Huang L, Reisenauer MR, Wu H, Chen L, Zhang Y, Xia Y, Zhang W · Genesis 2010
   PMID 21170927DOI
   ▸What this evidence cannot show

   Abstract read in full; full text not read. Murine germline knockout; no human data. 'No gross haematopoietic abnormality' is a steady-state observation - no transplant, no stress challenge, no competitive repopulation, and no cross onto a KMT2A fusion. The authors' own title says 'LIKELY not required'. The obvious paralogue MLLT10/AF10 was not deleted in the same animal, so redundancy is untested.

   ▸Evidence that argues against this (1)

   • The same laboratory subsequently reported a clear phenotype in the same mouse: Af17 deficiency lowers blood pressure by 11 mmHg, raises urine volume and sodium excretion, increases H3K79 dimethylation and reduces ENaC function. The knockout is therefore not phenotypically null; it is null in blood. Chen L et al. 2011
8. E_q69_ct2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   In MICE, Af17-deficient animals have lower blood pressure (by 11 mmHg), higher urine volume and increased sodium excretion despite mildly increased plasma aldosterone, and deletion of Af17 increases dimethylation of histone H3 K79 and reduces epithelial sodium channel function through decreased ENaC mRNA and protein, fewer active channels and lower open probability.

   Af17 deficiency increases sodium excretion and decreases blood pressure. · Chen L, Wu H, Pochynyuk OM, Reisenauer MR, Zhang Z, Huang L, Zaika OL, Mamenko M, Zhang W, Zhou Q, Liu M, Xia Y, Zhang W · J Am Soc Nephrol 2011
   PMID 21546577DOI
   ▸What this evidence cannot show

   Abstract read in full; full text not read. Murine, renal collecting duct, nothing haematopoietic and nothing about leukaemia. The relevance here is strictly as a cell-type answer and as a bound on the knockout's 'no phenotype' claim - it does not license any inference about AF17 in a leukaemic cell. High plasma aldosterone completely compensates for Af17 deficiency in sodium handling, so the phenotype is conditional on hormonal state.

   ▸Evidence that argues against this (1)

   • The haematopoietic negative remains the directly relevant result for this graph: the same allele produces no detectable blood phenotype. A renal requirement does not argue for a haematopoietic one, and this entry does not claim it does. Zhang Z et al. 2010

• related to →DPY30 — the smallest COMPASS subunit, and the only one that behaves like a selective dependency

  methodological: DotCom membership rests on unquantified MudPIT co-purification; quantitative stoichiometry reassigned DPY30's complex membership when it was applied to SET1/MLL

• part of →DOT1L — the target where excellent chemistry met an indirect interaction

  AF17 is a named DotCom subunit

• fuses with →KMT2A / MLL1 — the gene that breaks

  KMT2A::MLLT6, t(11;17)(q23;q21) — a fusion partner, but not in t(4;11)

• related to →MLLT10 (AF10) — DOT1L cofactor and KMT2A fusion partner

  the two non-YEATS MLLT proteins of DotCom

• related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)

  NOT an interactor: DotCom and the AF4/P-TEFb assembly are alternatives, and DOT1 does not bind AF9-AF4-P-TEFb complexes

• ← part ofCTNNB1 (beta-catenin) — the Wnt link inside DotCom, and the one piece of KMT2A biology that is not chromatin

  same complex (DotCom)

• ← co-occurs withKMT2A::SEPTIN9 (MLL-SEPT9 / MLL-MSF), t(11;17)(q23;q25)

  both arise from cytogenetic t(11;17) and are distinguished only molecularly

• ← related toNon-elongation KMT2A partners considered and not written: what the searches actually returned, partner by partner

  MLLT6 deferred; an entry already exists

• ← part ofKMT2A::MLLT6 (MLL-AF17), t(11;17)(q23;q12-q21) — the bridge case: a leucine zipper on a DOT1L-complex subunit

  AF17 leucine-zipper dimerisation motif, retained 3' of the fusion point; the N-terminal zinc fingers are lost

• ← fuses withKMT2A / MLL1 — the gene that breaks

  t(11;17)(q23;q12-q21); 27 patients, rank 11, adult-AML-weighted

• ← related toMLLT6 / AF17 — domain architecture, and where the leucine zipper actually is

  coordinate map for the same protein; gene-mllt6 holds the DotCom and dependency content and is not duplicated

• related to →DOT1L's localization record is a mouse kidney paper wearing a human label - and the one image of it in a leukaemia cell is a supporting figure a title search cannot see

  AF17 over-expression drives DOT1L/Dot1a into the cytoplasm by a leptomycin-B-sensitive route - a MEASURED cytoplasmic pool of DOT1L, against the 'by similarity' annotation that node is named for, but measured in renal epithelium

• related to →Two pools of one protein: the chromatin-bound XPO1 that matters here has never been separated biochemically from the soluble export machine

  the AF17-driven DOT1L export is leptomycin-B-sensitive and therefore CRM1/XPO1-dependent, putting a KMT2A fusion partner on the export axis

• related to →Bone marrow haematopoietic stem cell — the postnatal comparator, and why the same fusion behaves differently here

  EXPLICIT NEGATIVE: Af17-null mice show no gross haematopoietic abnormality at steady state, with no transplant, stress challenge or fusion cross performed (murine germline knockout; PMID 21170927)

• related to →DOT1L — the target where excellent chemistry met an indirect interaction

  the one tissue where Af17 loss has a measured phenotype - renal collecting duct - shows INCREASED H3K79 dimethylation, i.e. AF17 constrains the DOT1L mark in vivo outside blood (PMID 21546577)

• ← related toThe six unwritten DOT1L partners are not six findings - they are nine spokes of ONE GFP-DOT1L pulldown, and the source paper already graded them

  AF17 is the third A+ DotCom subunit recovered in the same pulldown, at 1:12.7

• ← related toTwo thirds of the curated interactome of this project's six core proteins is spoke-expanded - and the pairs this graph chose to write are significantly better than the ones it did not

  AF17's DotCom membership is 5 of 5 spoke-expanded with no shared structure - a genuine complex subunit whose contact with DOT1L has never been observed binary