Nothing connects ENL YEATS binding or ENL condensates to phosphorylation: zero records, and what that means for the phase-2 dossier
also called ENL condensate phosphorylation absence · YEATS phospho-gating absence · MLLT1 YEATS kinase
The ENL protein's reader pocket and its droplet behaviour are the two things a drug programme here would aim at, and we found no published link between either of them and phosphorylation. That matters before betting on the target: it means the question is open, not that the answer is no.
The same thing, for a scientist
This graph holds `target-phase2-enl-yeats-dossier`, a phase-2 START recommendation resting on two co-crystals (5J9S, 6HT1), 144 ChEMBL activity rows with a best reported Ki of 1.43 nM, and a leukaemia-conditional MEN1 co-dependency. The dispatch for this beat asked specifically whether YEATS-domain binding or ENL condensate behaviour is phosphorylation-gated. **The answer is that nobody has looked.** `(ENL OR MLLT1 OR YEATS1 OR LTG19)[tiab] AND (condensate OR condensates OR "phase separation" OR droplet OR droplets)[tiab] AND (phosphorylation OR phosphorylated OR phosphorylate OR kinase OR phosphosite)[tiab]` returns **0 records**. `("YEATS domain" OR "YEATS domains")[tiab] AND (phosphorylation OR phosphorylated OR kinase)[tiab]` returns 9, of which none addresses phospho-regulation of YEATS reader function; the closest is a paper in which the AF9 YEATS domain's TFIID interaction is controlled by acetylation and oligomerisation, with the phosphorylation acting elsewhere on the protein. This is an absence of testing and it is a genuine gap in an active dossier rather than a reassurance.
The queries and counts, 2026-09-19
| query | records | on-topic after title check |
|---|---|---|
(ENL OR MLLT1 OR YEATS1 OR LTG19)[tiab] AND (condensate OR condensates OR "phase separation" OR "phase-separation" OR droplet OR droplets)[tiab] AND (phosphorylation OR phosphorylated OR phosphorylate OR kinase OR phosphosite)[tiab] | 0 | 0 |
("YEATS domain" OR "YEATS domains")[tiab] AND (phosphorylation OR phosphorylated OR kinase)[tiab] | 9 | 0 |
(ENL OR MLLT1)[tiab] AND ("super elongation complex" OR SEC)[tiab] AND (phosphorylation OR phosphorylated OR kinase)[tiab] | 6 | 1 (the ATM/PRC1 review, which concerns the IDR not the YEATS domain) |
(MLLT1 OR YEATS1 OR LTG19)[tiab] AND (phosphorylation OR phosphorylated OR kinase)[tiab] | 9 | 2 (the ATM paper and its review) |
MLLT1[tiab] AND phosphosite[tiab] | 0 | 0 |
The nine 'YEATS domain' hits are: a bone phosphoproteomics paper about P70S6K and AFF4; an O-GlcNAcase SUMOylation paper; the AF9 oligomerisation paper; a heart-failure bioinformatics paper; a ZAK-alpha kinase structure paper; a prokaryotic immune-systems survey; a YEATS4/GAS41 gastric cancer paper; a lipomatous-tumour copy-number paper; and a 2010 review of YEATS domain structure. None tests whether phosphorylation changes YEATS-domain ligand binding.
What the atlas says about the YEATS domain specifically
A second, independent bound comes from the phosphosite data rather than the literature. The ENL YEATS domain spans residues 1-138. Of the 23 observed phosphosites on Q03111, the most N-terminal is S155 - so not one observed phosphosite lies inside the ENL YEATS domain 2.
The same holds for AF9: its YEATS domain is 1-138 and its most N-terminal observed site is S288 2.
That is a real constraint on the hypothesis the dispatch raised. Direct phospho-gating of the reader pocket by a phosphate within the domain has no proteomic support in either paralogue. It does not exclude allosteric gating from the IDR, which is where all the observed sites are, nor gating of the acetyl-lysine ligand rather than the reader - but the simplest version of the hypothesis is disfavoured before any experiment.
What this means for the dossier - and it is not 'no'
Three things follow for target-phase2-enl-yeats-dossier.
First, the G2 biology gate is unaffected. That gate asks whether removing ENL collapses the fusion output programme, and phospho-regulation is orthogonal to it.
Second, an unknown appears in the G1 feasibility gate. If YEATS-pocket occupancy is modulated by a phosphorylation somewhere on ENL, then cellular target engagement measured by thermal shift could vary with the signalling state of the cell - between cell lines, between treatment conditions, and between a cycling culture and a patient blast. Nobody knows whether it does, and the dossier does not list it as a variable.
Third, the condensate leg needs the standing caveat attached. The dossier's neighbouring entries treat ENL condensates as a mechanism. This graph's rule is that hexanediol sensitivity and classical FRAP cannot distinguish phase separation from binding to spatially clustered sites 3, and hexanediol itself condenses chromatin 4. The cancer-ENL condensate literature is better than average on this point - it works at endogenous expression and reports partial rather than complete FRAP recovery, and explicitly states that expression beyond endogenous levels yields non-functional condensates 5 - but the dossier should say which model it means.
And the cerebellum observation this graph holds, MLLT1 at 141.7 TPM in Brain_Cerebellum against 40.3 in whole blood, remains entirely unexplained: (MLLT1 OR ENL)[tiab] AND (cerebellum OR cerebellar)[tiab] returns 2 records, neither about ENL function there. A phospho-regulatory hypothesis for that tissue difference has no evidence in either direction.
| status | run |
|---|---|
| gene | gene-mllt1 |
| uniprot | Q03111 |
| records condensate x phospho | 0 |
| records yeats domain x phospho | 9 |
| records yeats domain x phospho on topic | 0 |
| records mllt1 x phosphosite | 0 |
| records mllt1 x cerebellum | 2 |
| observed phosphosites in yeats domain enl | 0 |
| observed phosphosites in yeats domain af9 | 0 |
| absence class | never-measured |
| bearing on dossier | target-phase2-enl-yeats-dossier: G2 unaffected; G1 acquires an unmeasured variable (does signalling state change pocket occupancy?); the condensate leg needs an explicit model statement |
| what would overturn this | Any study showing a phosphorylation that changes YEATS-domain ligand affinity, or that gates ENL condensate formation; or a phosphosite mapped inside residues 1-138 of either paralogue. |
| result summary | Zero published records linking ENL YEATS binding or ENL condensates to phosphorylation, and zero observed phosphosites inside the YEATS domain of either ENL or AF9. Open question, not a closed one. |
| gene symbol | MLLT1 |
| absence class previous freetext | absence of testing |
| absence class basis | Zero records; the assay pairing condensate behaviour with phosphorylation has not been performed on ENL. |
| absence class assigned | 2026-09-21 (cycle 47, king) |
yeats domain boundaries
{
"start": 1,
"end": 138
}most n terminal observed site
{
"MLLT1": 155,
"MLLT3": 288
}- The ENL YEATS pocket is a phase-2 START in this graph, and a target whose ligand-binding might vary with cellular signalling state would give inconsistent engagement data across cell lines for reasons nobody would attribute correctly. This is the cheap variable to eliminate before, not after, the biology gate.
- A condensate-disruption strategy built on assays that cannot distinguish phase separation from clustered binding risks optimising a molecule against the wrong physical model. The distinguishing assay exists and is published.
- Because all observed ENL phosphorylation is in the IDR and the IDR's charge is what the condensate mechanism turns on, a kinase inhibitor could plausibly modulate condensates as an off-target consequence of unrelated signalling inhibition - a confound for any condensate experiment run in cells under drug treatment.
- Does the signalling state of a cell change ENL YEATS-pocket occupancy by SGC-iMLLT? A cellular thermal shift assay run in the presence and absence of a kinase inhibitor panel would answer it and belongs in the dossier's G1 gate.
- Are ENL condensates phospho-gated? The prediction is testable with the existing endogenous-level Halo-ENL system and a kinase inhibitor.
- All 23 observed ENL phosphosites lie in the IDR, which is one of the two oppositely charged disordered regions the condensate mechanism requires. Does phosphorylating that region - which adds negative charge - modulate the charge-based multivalency the mechanism depends on? That is a specific, mechanistically motivated experiment nobody has run.
- MLLT1 peaks at 141.7 TPM in cerebellum with no published function there and two records in total. Is the difference post-translational?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
PubMed returns zero records for any combination of ENL/MLLT1/YEATS1/LTG19 with condensate or phase-separation terms and phosphorylation or kinase terms, and zero for MLLT1 combined with 'phosphosite'; the nine records combining 'YEATS domain' with phosphorylation terms contain none that tests phospho-regulation of YEATS reader function.
PubMed esearch, five stated queries run 2026-09-19 with title checking of all returned records · NCBI E-utilities 2026What this evidence cannot show
A counted absence over stated [tiab] queries. It cannot see work described only in a supplementary figure, a preprint not indexed in PubMed, or a paper that reports the result without using 'phosphorylation', 'kinase', 'condensate' or 'phase separation' in title or abstract. 'ENL' is a badly behaved search token (see ana-enl-token-collision-alias-precision) but here it can only inflate the count, so a zero is safe against that particular failure.
Evidence that argues against this (1)
- A zero count over condensate-plus-phosphorylation terms does not mean phospho-regulation of ENL condensates has been excluded; it means the intersection has not been published. The condensate literature on ENL is recent, active and mechanistically detailed, and a study of this exact question could appear at any time - so this absence has a short half-life. Song L et al. 2022
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Not one of the 23 observed phosphosites on ENL (Q03111) lies within the YEATS domain (residues 1-138): the most N-terminal observed site is S155. The same holds for AF9 (P42568), whose YEATS domain is 1-138 and whose most N-terminal observed site is S288.
Proteins API proteomics-ptm Q03111 and P42568, fetched 2026-09-19; positions compared against the YEATS domain boundaries recorded in this graph · EMBL-EBI 2026What this evidence cannot show
An argument from proteomic coverage, which is exactly the kind of argument this beat warns against elsewhere: a folded, structured domain yields different tryptic peptides from a disordered region and can be systematically under-sampled. The AF9 S395 precedent is directly relevant - a functionally decisive site that appears in no proteomic resource. So this constrains the hypothesis without excluding it.
Evidence that argues against this (1)
- The controlling counter-example is in this beat: AF9 S395 has a kinase, a mutant phenotype and a mechanism, and appears in neither the atlas nor UniProt's curated table. Absence from phosphoproteomics is a weak argument against a damage- or signal-induced site, and the YEATS domain has never been assayed under a stimulus. Talukdar P et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
1,6-hexanediol sensitivity and classical FRAP recovery both fail to distinguish liquid-liquid phase separation from binding to spatially clustered sites; only model-free calibrated half-bleaching measuring internal mixing discriminates the two.
Detecting and quantifying liquid-liquid phase separation in living cells by model-free calibrated half-bleaching. · Muzzopappa F, et al. · Nat Commun 2022What this evidence cannot show
A methods paper establishing what the assays cannot do. It does not show that any specific published condensate is misassigned.
Evidence that argues against this (1)
- Clustered binding is a substantive rival model rather than a debunking, and in a chromatin context arguably the more interesting one, because clustered binding predicts sequence-specific targeting where phase separation predicts bulk enrichment. Treating the methodological caveat as evidence against condensates inverts its meaning. Song L et al. 2022
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
1,6-hexanediol rapidly immobilises and condenses chromatin in living human cells, an action the authors state is distinct from dissolving a droplet and which they instruct readers to use in reinterpreting chromatin-associated droplet results.
1,6-hexanediol rapidly immobilizes and condenses chromatin in living human cells. · Itoh Y, et al. · Life Sci Alliance 2021What this evidence cannot show
Human cell lines; an argument about the reagent, not about ENL.
Evidence that argues against this (1)
- The reagent critique does not reach studies that established their condensates by other means, and the principal ENL condensate study builds its case on mutagenesis, expression-level titration and chromatin-bound imaging rather than on hexanediol alone. Song L et al. 2022
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Cancer-associated ENL YEATS hotspot mutations trigger condensates at native genomic targets through multivalent homotypic and heterotypic interactions requiring mutation-induced structural changes in the YEATS domain, ENL's two oppositely charged disordered regions and incorporation of extrinsic elongation factors; at endogenous levels ENL mutants form submicron chromatin-bound condensates with partial FRAP recovery, while expression beyond the endogenous level gives larger, highly dynamic and non-functional condensates.
Hotspot mutations in the structured ENL YEATS domain link aberrant transcriptional condensates and cancer. · Song L, et al. · Mol Cell 2022What this evidence cannot show
Full text read via efetch db=pmc (PMC10071517) after fullTextXML returned HTTP 500. The system is ENL YEATS hotspot mutants, which occur in Wilms tumour and a subset of AML, not KMT2A::MLLT1 fusions; the condensate mechanism is not shown for the fusion. FRAP is the principal dynamics assay, which is subject to the discrimination caveat above.
Evidence that argues against this (1)
- A structural review of the same protein family attributes ENL's transcriptional role to phase separation enhancing SEC recruitment without reproducing the expression-dependence distinction, which shows how quickly the careful version of this result is flattened in secondary literature. Kabra A et al. 2022
generalises the ENL YEATS observation across ten proteins and 46 sites with no exception
re-ran the condensate-phospho query this cycle: (ENL OR MLLT1) AND condensate AND phospho* still returns 0 with a clean querytranslation
the phospho-condensate zero re-measured this cycle, with the ATM SQ-site literature named as the reason the zero is about the intersection
phospho-gating of the pocket is untested; adds a variable to G1
no observed phosphosite inside the ENL YEATS domain
the retained reader has no known kinase input
cerebellar MLLT1 remains unexplained; 2 records
ENL YEATS hotspot condensates, no phospho link
why the zero counts survive the noisy token
counted absence 1
SURVIVES at full text: 40 records read, 11 proceedings, and the three primary YEATS condensate papers contain no phospho arm
SURVIVES an ENL YEATS agent-name re-query (SR-0813, TDI-11055, SGC-iMLLT)