AKAP95 directly enhances H3K4 mono-, di- and tri-methylation - by the KMT2B complex, and ONLY by the KMT2B complex; the paper's own discussion says the MLL1/KMT2A case was never tested
also called AKAP95 enhances MLL2C HMT activity · AKAP95 H3K4 methylation modulator
AKAP95 makes one particular H3K4-methylating machine work better - the one built around KMT2B. People often quote this as if it applied to KMT2A, the gene that matters in this leukaemia, but that experiment was never done and the authors said so.
The same thing, for a scientist
Recombinant FLAG-AKAP95 purified from Sf9 cells, added to an in vitro HMT reaction on chromatin assembled from recombinant histone octamers and plasmid DNA, significantly increases **MLL2C**-mediated H3 methylation; immunoblot resolves the effect across **H3K4me1, me2 and me3** [E1]. Two negative controls in the same figure make it an effect on the enzyme rather than an artefact: purified AKAP95 has **no methyltransferase activity of its own**, and it **does not associate with free histone H3**, so neither a contaminating MTase nor substrate sequestration explains the result. **AKAP95(101-692) does not do it**, tying the effect to the same N-terminal region that binds DPY30 [E1]. **The scope limit is the whole point of this entry.** 'MLL2' in this paper is **KMT2B** (the authors cloned it from KIAA0304, gene ID 9757); UniProt calls the same gene MLL4/WBP7. The paper's discussion states: *'It will also be interesting to examine whether AKAP95 directly enhances H3K4 methylation mediated by other MLL complexes'* [E2]. **No enzymatic experiment with MLL1/KMT2A exists.**
Exactly what supports what - three different MLL1 statements, only one of them enzymatic
The paper makes three claims that a fast reader merges into one. They are not the same.
| claim | MLL1/KMT2A? | assay | strength |
|---|---|---|---|
| AKAP95 co-elutes with MLL1 and MLL2 complexes on Superose 6, but not with free or dimerised DPY30 | yes | gel filtration - no bait, symmetric | association at complex level; cannot distinguish a stable subunit from a co-sized assembly |
| Ectopic AKAP95 synergises with MLL1 or MLL2 on a chromosomal Gal4-luciferase reporter | yes | cell-based reporter in G-293T | functional synergy, but the readout is transcription, three steps downstream of methylation |
| Recombinant AKAP95 enhances H3K4me1/me2/me3 on chromatin | NO - MLL2C only | purified in vitro HMT | the only direct enzymatic result, and it is not the MLL1 complex |
So 'AKAP95 enhances MLL methyltransferase activity' is true of KMT2B and unmeasured for KMT2A. The abstract's phrase 'directly enhances their methyltransferase activity' uses a plural that the data do not cover, and UniProt's curators read it correctly: their SUBUNIT line scopes the DPY30-mediated association to 'at least the MLL4/WBP7 HMT complex', which is KMT2B 3.
And the authors say the complexes are not interchangeable. Their MLL2C carried no detectable HAT, whereas the previously purified WDR5-MLL1 complex carried MOF - a difference they attribute either to tagging strategy or to genuine MLL1/MLL2 sequence differences 2. That is an argument against transferring an activity measurement from one to the other.
Why this does not transfer to KMT2A::AFF1 even if it were shown for MLL1
KMT2A::AFF1 loses the SET domain. The fusion has no H3K4 methyltransferase activity for AKAP95 to enhance. A rate modifier of an enzyme is irrelevant to a protein that is not that enzyme.
The most this line of evidence could ever support in a KMT2A-rearranged cell is a modifier effect on the wild-type allele - and every KMT2A-rearranged blast retains one wild-type KMT2A, which this graph already treats as a live variable (concept-a-wdr5-result-is-not-automatically-a-kmt2a-axis-result makes the same distinction for WDR5).
The route that does transfer is a different one entirely - IDR co-condensation with the retained N-terminal fragment, which needs no catalysis at all (interface-akap8-kmt2a-n-terminal-fragment). Keeping these two apart is the single most useful thing this beat can hand the graph, because the enzymatic story is the memorable one and it is the one that does not apply.
| enzyme | gene-kmt2b |
|---|---|
| modifier | gene-akap8 |
| mark | H3K4me1, H3K4me2, H3K4me3 |
| direction | positive - increases deposition |
| system | purified in vitro HMT on recombinant chromatin (recombinant octamers + plasmid DNA), 3H-SAM autoradiography plus state-specific immunoblot |
| shown for fusion | NO - nobody has looked in any KMT2A fusion, and the mechanism is a priori inapplicable because KMT2A::AFF1 loses the SET domain. The most it could bear on is the retained wild-type KMT2A allele. |
| shown for kmt2a | NO - no HMT assay with an MLL1/KMT2A complex exists in this or any follow-up paper found by this beat. The MLL1 statements in the paper are gel-filtration co-elution and a cell-based reporter synergy. |
measurements
[
{
"quantity": "fold_change",
"subject": "MLL2C H3K4 methylation on recombinant chromatin, +AKAP95 vs -AKAP95",
"value": null,
"units": "fold",
"n": null,
"universe": "PMID 23995757 Fig 6B/6C; no densitometry, no statistic and no fold-change is reported anywhere in the text",
"null": null,
"note": "RECORDED AS NULL DELIBERATELY. The headline effect of this entry has never been quantified. Do not let a later reader infer a number."
},
{
"quantity": "mass",
"subject": "recombinant AKAP95 per HMT reaction",
"value": 0.5,
"units": "ug",
"n": null,
"universe": "PMID 23995757 Methods",
"null": null,
"note": "against 250-350 ng chromatin DNA and a similar mass of octamer - i.e. AKAP95 is in large excess over the template"
}
]- If AKAP95 is a rate modifier of H3K4 methylation by the SET1/MLL family generally, an AKAP95 inhibitor hits wild-type KMT2A, KMT2B, KMT2C/D and both SET1 complexes at once. There is no equivalent of the menin-interface argument that gives menin inhibitors their window.
- A stimulation demonstrated only in a purified reaction with the modifier in large excess over template is the class of result that fails to reproduce in a nucleus, where AKAP95 is one of thousands of chromatin proteins and is not free.
- This entry exists partly to stop the graph asserting an enzymatic KMT2A effect. If a downstream analysis pulls 'AKAP95 enhances MLL methyltransferase activity' out of a summary field without the KMT2B scope, it will licence a wet experiment that has no mechanistic basis in a SET-less fusion.
- Does AKAP95 enhance the MLL1/KMT2A complex's methyltransferase activity? The reagents to answer it existed in 2013 in the same laboratory and the authors flagged it as the obvious next experiment; thirteen years later no published answer exists.
- Is the enhancement DPY30-dependent? The inference that it works 'likely through binding to DPY30' is the authors' own hedge. A DPY30-free reconstitution, or the AKAP95 construct that fails to bind DPY30, would test it directly - and the second reagent already exists.
- Does AKAP95 loss change H3K4me3 at HOXA9 and MEIS1 in a KMT2A-rearranged cell? The 2026 paper measured H3K27ac at Meis1 and Pbx3 in mouse pre-leukaemia and H3K4me3 only at the differentiation markers, so this specific question is still open in the disease.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Recombinant FLAG-HA-AKAP95, purified to near homogeneity from baculovirus-infected Sf9 cells and used at about 0.5 ug per reaction, significantly enhances MLL2-complex-mediated H3 methylation on chromatin assembled from recombinant histone octamers and 250-350 ng plasmid DNA, with H3K4 mono-, di- and tri-methylation all enhanced by immunoblot; AKAP95(101-692) at comparable amount does not. Purified full-length and mutant AKAP95 possess no methyltransferase activity of their own and do not associate with free histone H3.
Regulation of transcription by the MLL2 complex and MLL complex-associated AKAP95. · Jiang H, Lu X, Shimada M, Dou Y, Tang Z, Roeder RG · Nat Struct Mol Biol 2013What this evidence cannot show
Figure 6B-D; full text and Methods read, figure panels not inspected, so 'significantly enhanced' is the text's word and no fold-change or statistic is given anywhere in the text. The reaction is a purified system with one enzyme complex, so it says nothing about whether the effect survives in a nucleus. The expected_absent_terms are the load-bearing part of this record: no HMT assay with an MLL1/KMT2A complex appears anywhere in the paper.
Evidence that argues against this (1)
- The stimulation is by an amount the paper never quantifies. In a project that requires typed measurements, an unquantified 'significant enhancement' from a densitometry-free autoradiograph is the weakest form a positive in vitro result can take, and it has not been independently reproduced in the thirteen years since. Jiang H et al. 2013
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The paper's Discussion states that 'As DPY30 is shared among all MLL complexes, it is possible, but has not been rigorously demonstrated, that AKAP95 may also be commonly associated with all MLL complexes' and that 'It will also be interesting to examine whether AKAP95 directly enhances H3K4 methylation mediated by other MLL complexes.' It also notes that its MLL2 complex contained no detectable HAT while a previously purified WDR5-MLL1 complex contained MOF, attributing the difference either to tagging and purification strategy or to MLL1/MLL2 sequence variation.
Regulation of transcription by the MLL2 complex and MLL complex-associated AKAP95. · Jiang H, Lu X, Shimada M, Dou Y, Tang Z, Roeder RG · Nat Struct Mol Biol 2013What this evidence cannot show
Direct quotation from the Discussion of the full text. This is the authors' own scope statement and it is stronger evidence for the limit than any inference from the figures.
Evidence that argues against this (1)
- An author's hedge in a Discussion records what they chose to claim, not what they measured, and this entry quotes the cautious sentence while the same paper's ABSTRACT says AKAP95 'directly enhances their methyltransferase activity' with a plural covering the DPY30-MLL complexes. Both sentences are in the paper. The selection made here is deliberate and defensible - the figures support only the KMT2B assay - but a reader should know the confident phrasing exists, because it is the one that propagates. Jiang H et al. 2013
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt's SUBUNIT comment for O43823 scopes the DPY30-mediated association to 'at least the MLL4/WBP7 HMT complex', citing PubMed 23995757. In UniProt's naming MLL4/WBP7 is KMT2B (Q9UMN6, 2715 aa) - the same gene the source paper calls MLL2 and cloned from KIAA0304.
UniProtKB entry O43823 (AKAP8_HUMAN) SUBUNIT comment, and O43823/Q9UMN6 cross-check, retrieved live 2026-09-24 by q166 · UniProt Consortium 2026What this evidence cannot show
This project's `data/symbol-collisions.json` records MLL2 and MLL4 as aliases shared by KMT2B (Q9UMN6, 2715 aa) and KMT2D (O14686, 5537 aa), with each protein name carrying the other's number, and length as the only discriminator. The identification of the source paper's 'MLL2' as KMT2B is not made from the alias: the paper states 'MLL2 (Kmt2b, gene ID 9757)' and names KIAA0304 as the cDNA, and KIAA0304 is a KMT2B synonym in the UniProt record. See `concept-mll2-and-mll4-each-name-two-genes`.
Evidence that argues against this (1)
- A curator's alias choice is not independent corroboration of gene identity in exactly this alias family. MLL2 and MLL4 are shared aliases of KMT2B and KMT2D, with each protein name carrying the other's number, so a UniProt line reading 'MLL4/WBP7' inherits whatever that curator took the alias to mean. The gene identification in this entry does not rest on it and must not be allowed to: it rests on the paper's own 'MLL2 (Kmt2b, gene ID 9757)' and on KIAA0304 being a KMT2B synonym. this project 2026
the enzymatic arm of AKAP95's function, and the arm that does NOT transfer to a SET-less fusion
AKAP95 increases MLL2C H3K4 me1/me2/me3 on recombinant chromatin; the effect is on the enzyme, not the substrate
- related to →MLL2 and MLL4 each name two different genes, and the ambiguity is in the papers, not in the databases
the paper's 'MLL2' is KMT2B and UniProt's 'MLL4/WBP7' for the same gene; the alias is why this result gets quoted as a KMT2A result
AKAP95(101-692) neither binds DPY30 well nor enhances methylation - the authors' own inference that the effect runs through DPY30
claim 5 scored CONFIRMED but scope-limited to KMT2B by the authors' own Discussion
Route A, the enzymatic route, which cannot operate on a SET-less fusion