AKT and ATM push ELL between two complexes: K355 acetylation sends it to SEC, ATM phosphorylation sends it to LEC
also called ELL K355 · ELL LEC SEC partition · AKT p300 ELL · Sin3A HDAC1 ELL
ELL belongs to two different machines - one that reads out growth genes and one that reads out small housekeeping RNAs. Which machine it joins is decided by a chemical tag, and two different kinases push the tag in opposite directions: a growth signal sends ELL to the growth machine, and a DNA-damage signal sends it to the other one.
The same thing, for a scientist
ELL partitions between the Super Elongation Complex and the Little Elongation Complex, and the partition is set by the acetylation state of Lysine 355. p300-mediated K355 acetylation favours ELL monomerisation, ELL-SEC assembly and transcription of proliferation-related mRNA genes; HDAC1-mediated deacetylation favours dimerisation, ELL-LEC assembly and snRNA transcription. Two kinase inputs set the direction: **mitogen treatment enhances AKT signalling-dependent p300-mediated K355 acetylation**, driving ELL into the SEC; and **genotoxic stress causes ATM-mediated ELL phosphorylation-dependent deacetylation of K355 by the Sin3A-HDAC1 complex**, driving ELL into the LEC. K355 is retained in MLL-ELL (which contains ELL 46-621), as are the ATM sites. The consequence for this graph is specific: the fusion's SEC engagement - the mechanism by which KMT2A::ELL is thought to work - is downstream of a growth-factor kinase and a damage kinase pulling in opposite directions on a retained residue.
The two-kinase partition
| input | route | ELL state | complex | output |
|---|---|---|---|---|
| mitogen / AKT signalling | p300 acetylates K355 | monomer | SEC | proliferation-related mRNA |
| genotoxic stress / ATM | ATM phosphorylates ELL, enabling Sin3A-HDAC1 deacetylation of K355 | dimer | LEC | snRNA |
Both directions from the same primary report 1. The ATM arm connects this entry to mechanism-atm-ell-t551-s561-s589-eaf1-switch: the same kinase, the same protein, and in the earlier paper the same consequence of leaving the SEC. The two papers come from one laboratory and describe two mechanisms by which ATM removes ELL from the SEC - increased EAF1-driven self-association in one, K355 deacetylation and LEC assembly in the other. Whether these are one mechanism or two is not established and the entries should be read together rather than added up.
Retained - and this is the retained-half argument in its strongest form
K355 is within ELL 46-621 and is therefore in the fusion. The authors' own mapping narrows the HDAC1-responsive region to ELL residues 45-373, and they use a GST-ELL (61-621) construct in the deacetylation assay - both firmly inside the retained portion 1.
So KMT2A::ELL should carry an AKT-responsive acetyl switch that controls whether it engages SEC. That is an actionable, untested proposition: AKT and PI3K inhibitors are widely available, and this graph already holds mechanism-gsk3-dependency-kmt2a-r-leukaemia and FLT3-signalling entries for KMT2A-rearranged disease, so the growth-factor axis is present but has never been connected to the fusion protein itself.
| writer | p300 |
|---|---|
| eraser | HDAC1 / Sin3A-HDAC1 |
| substrate | gene-ell |
| uniprot | P55199 |
| site | K355 (acetylation, kinase-gated in both directions) |
| gated contact | partition between Super Elongation Complex and Little Elongation Complex |
| direction acetyl | K355ac -> monomer -> SEC -> proliferation mRNA |
| direction deacetyl | K355 deacetylated -> dimer -> LEC -> snRNA |
| mapped in | wild-type ELL, mammalian cell lines |
| mapped in fusion | false |
| retained in fusion | true |
| retained basis | MLL-ELL contains ELL aa 46-621; K355 is inside that, as is the authors' 45-373 HDAC1-responsive region |
| gene symbol | ELL |
kinases
[ "AKT (via p300)", "ATM (via Sin3A-HDAC1)" ]
inhibitors used in primary
[ "MS275 (HDAC1)", "RGFP966 (HDAC3)", "C646 (p300)" ]
same laboratory as
[ "mechanism-atm-ell-t551-s561-s589-eaf1-switch", "interface-ell-eaf1-phospho-gated-condensate" ]
- PI3K and AKT inhibitors are already used in haematological malignancy. On this mechanism an AKT inhibitor would deacetylate K355 and push ELL - and by inference the fusion - out of the SEC, which sounds therapeutic; but the same axis controls normal proliferative transcription in every cell, so the selectivity argument is entirely absent.
- HDAC inhibitors would do the opposite: preserving K355 acetylation would favour ELL-SEC assembly and, on this model, favour the fusion's oncogenic complex. HDAC inhibitors are trialled in KMT2A-rearranged AML, and this is a mechanistic reason they might antagonise rather than help in the ELL subgroup specifically.
- The whole partition model is one laboratory across three papers, with many single-replicate panels and no leukaemic system. It is a hypothesis-generating architecture, not a platform.
- Does KMT2A::ELL carry K355 acetylation, and does a PI3K/AKT inhibitor move the fusion out of the SEC? This is a co-IP and a western, with drugs that are on the shelf.
- Are the ELL/ATM/EAF1 self-association mechanism and the ELL/ATM/K355 LEC mechanism the same event described twice, or two? Both come from one laboratory and both end with ELL leaving the SEC.
- KMT2A::ELL patients are enriched in secondary and therapy-related AML in this graph's own patient entries. Does prior genotoxic therapy leave a durable mark on this partition?
- Does MLL-ELL enter the LEC at all? If the fusion can be pushed into an snRNA-transcribing complex, that is a mechanism of inactivation with no drug currently aimed at it.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
p300-mediated acetylation of ELL at Lysine 355 favours ELL monomerisation and ELL-SEC assembly with transcription of proliferation-related mRNA genes, while HDAC1-mediated deacetylation favours dimerisation and ELL-LEC assembly with snRNA transcription; mitogen treatment enhances AKT-signalling-dependent p300-mediated K355 acetylation, and genotoxic stress causes ATM-mediated ELL-phosphorylation-dependent deacetylation of K355 by the Sin3A-HDAC1 complex.
Dynamic partitioning of a critical elongation factor between LEC and SEC regulates cellular snRNA and proliferation-related mRNA transcription. · Nandy A, et al. · Nat Commun 2026What this evidence cannot show
Full text read via Europe PMC fullTextXML on 2026-09-19. A 2026 paper from the same laboratory as the ELL/ATM and ELL/EAF1 reports, so the three are not independent observations. Many panels are stated as n=1 or n=2 replicates, including the K355Ac antibody validation. The K355 assignment rests on a mutant series informed by a previous high-throughput acetylation study. Mammalian cell lines; no leukaemic system and no MLL-ELL fusion protein. The AKT link is pharmacological and correlative rather than a direct demonstration that AKT phosphorylates p300 to this end.
Evidence that argues against this (1)
- The founding characterisation of the Little Elongation Complex establishes it as an snRNA-specific machine built on ICE1 and ICE2 with ELL as a shared subunit, and does not describe a regulated partition between LEC and SEC. A dynamic acetylation-controlled exchange between the two is a substantially newer and less replicated model than the complexes themselves. Smith ER et al. 2011
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
K355 lies within the ELL sequence retained in MLL-ELL, which contains ELL amino acids 46 to 621; the authors' own mapping of the HDAC1-responsive region to ELL 45-373 and their use of a GST-ELL(61-621) construct both fall inside the retained portion.
The elongation domain of ELL is dispensable but its ELL-associated factor 1 interaction domain is essential for MLL-ELL-induced leukemogenesis. · Luo RT, et al. · Mol Cell Biol 2001What this evidence cannot show
The 46-621 boundary is taken from this graph's subtraction-ell-n-terminal-eaf-site entry, which resolved it from this paper's full text in a previous cycle; this session did not re-read it. The inference that K355 is therefore functional in the fusion is a sequence argument only - no one has expressed MLL-ELL and measured K355 acetylation.
Evidence that argues against this (1)
- Sequence retention is not functional retention: the same graph entry records that the retained ELL elongation-stimulatory region is dispensable for MLL-ELL leukaemogenesis despite being present in the fusion. A retained residue can be present and irrelevant. Simone F et al. 2003
K355 acetylation sets LEC/SEC partition
p300 is the K355 writer
- related to →KMT2A::ELL (MLL-ELL), t(11;19)(q23;p13.1)
K355 retained in MLL-ELL
same kinase, same protein, two published routes out of the SEC
ATM phospho enables Sin3A-HDAC1 deacetylation of ELL K355
relation 4 in the ledger