The only cellular proposal for how the SEC gets P-TEFb out of the reservoir is a condensate, and the two assays it rests on cannot tell that model from its rival
also called SEC phase separation · ENL droplets · P-TEFb compartmentalisation
One published idea is that the elongation scaffold clumps together into liquid droplets and sucks the enzyme out of the off-switch particle by sheer local concentration. It is a good idea and the experiments used to support it cannot distinguish it from a much simpler explanation, so this entry records both.
The same thing, for a scientist
Guo and colleagues report that SEC components compartmentalise and concentrate P-TEFb by liquid-liquid phase separation, drawing it away from the soluble inactive HEXIM1-containing complex; ENL or its intrinsically disordered region alone initiates droplet formation, AFF4 fluidises the droplets, droplets form within minutes of serum exposure and are required for rapid transcriptional induction, and **fusion of ENL with MLL boosts SEC phase separation** [E1]. If correct this is the cellular answer to the question Tat answers virally: how P-TEFb leaves the reservoir without a protein that binds 7SK RNA. It is also, as written, the strongest mechanistic link in this beat between a KMT2A fusion and the reservoir. Three published methodological results bound how much weight it can carry. 1,6-hexanediol sensitivity and classical FRAP recovery **both fail to distinguish phase separation from binding to spatially clustered sites**, and only model-free calibrated half-bleaching measuring internal mixing discriminates them [E2]. 1,6-hexanediol additionally immobilises and condenses chromatin in living human cells, doing something close to the opposite of what the assay assumes [E3]. And at physiological expression, the low-complexity-domain interactions that control transcription form **dynamic hubs without detectable phase separation** [E4]. The rival model - clustered binding sites rather than a phase - is not a way of dismissing the result; it is the more specific and arguably more interesting model, because clustered binding predicts sequence-directed targeting where a phase predicts bulk enrichment.
What the claim would mean for KMT2A::AFF1 if it holds
The Guo result is about MLL-ENL, not KMT2A::AFF1, and the boosted phase separation is attributed to ENL's intrinsically disordered region 1. Transferring it to KMT2A::AFF1 requires an extra step that the paper does not take: AFF1 is a different protein, the fusion retains AFF1 from residue 347 onward including the pSER region and CHD, and no equivalent droplet experiment has been reported for a KMT2A::AFF1 fusion.
It also interacts with this beat's central structural finding. The fusion has no cyclin T1-binding sequence (see mechanism-kmt2a-aff1-cannot-be-sequestered-in-the-7sk-snrnp), so if the fusion does concentrate P-TEFb it must do so by concentrating a partner that binds cyclin T1 - AFF4 or wild-type AFF1 - rather than by holding P-TEFb itself. A condensate model is actually the most natural way for an in-trans recruiter to work, which makes this the one place where the two findings reinforce each other rather than competing.
What would settle it
Calibrated half-bleaching on endogenous, tagged ENL or AFF4 at physiological expression, measuring internal mixing rather than recovery 2; and a demonstration that P-TEFb is depleted from the soluble HEXIM1-bound fraction when droplets form, which is a biochemical fractionation rather than an imaging result. Neither has been done in a KMT2A-rearranged cell.
This graph already holds localization-condensate-evidence-standards and localization-condensate-strongest-positive-case; this entry is the reservoir-facing instance of the same problem and should be read alongside them.
| mechanism class | proposed physical mechanism, model-underdetermined |
|---|---|
| discriminating assay | model-free calibrated half-bleaching measuring internal mixing (PMID 36526633) |
| assays used in the source | droplet imaging in overexpression systems |
| fusion studied | MLL-ENL (KMT2A::MLLT1), not KMT2A::AFF1 |
rival models
[ "liquid-liquid phase separation of SEC components concentrating P-TEFb", "binding to spatially clustered sites with no phase" ]
related entries
[ "localization-condensate-evidence-standards", "localization-condensate-strongest-positive-case", "hyp-aff1-condensates-dispensable-for-kmt2a-fusion-function", "mechanism-sec-elongation-hijack" ]
- If the mechanism is clustered binding rather than a phase, then agents designed to dissolve condensates will not work, and 1,6-hexanediol data supporting them is uninterpretable.
- A condensate-dissolving strategy targets a physical property shared by many nuclear assemblies, so selectivity for the fusion's assembly is not obviously achievable.
- Does calibrated half-bleaching on endogenously tagged ENL or AFF4 show internal mixing, or clustered binding?
- Does droplet formation deplete P-TEFb from the soluble HEXIM1-bound fraction, measured biochemically rather than by imaging?
- Does KMT2A::AFF1 boost SEC phase separation the way MLL-ENL does, given that it contributes AFF1's pSER and CHD rather than ENL's intrinsically disordered region?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
SEC components compartmentalise and concentrate P-TEFb via liquid-liquid phase separation away from the soluble inactive HEXIM1-containing P-TEFb complex; ENL or its intrinsically disordered region is sufficient to initiate SEC droplet formation, AFF4 fluidises the droplets, droplets form rapidly on serum exposure and are required for rapid transcriptional induction, and fusion of ENL with MLL boosts SEC phase separation.
ENL initiates multivalent phase separation of the super elongation complex (SEC) in controlling rapid transcriptional activation. · Guo C, Che Z, Yue J, Xie P, Hao S, Xie W, Luo Z, Lin C · Sci Adv 2020What this evidence cannot show
Imaging-led, largely in overexpression systems, with droplet assays as the primary readout. The fusion studied is MLL-ENL, not KMT2A::AFF1. The confidence is set at moderate because the headline claim is a model that the assays used cannot separate from a rival model, not because the observations are doubted.
Evidence that argues against this (2)
- 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 that measures internal mixing discriminates the two. Muzzopappa F et al. 2022
- At physiological expression, low-complexity-domain interactions controlling transcription form dynamic hubs with no detectable phase separation, so punctate appearance under overexpression is weak evidence about the endogenous protein. Chong S et al. 2018
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The two commonest condensate assays - 1,6-hexanediol sensitivity and classical FRAP recovery - cannot distinguish liquid-liquid phase separation from binding to spatially clustered sites; model-free calibrated half-bleaching, which measures internal mixing, does discriminate them.
Detecting and quantifying liquid-liquid phase separation in living cells by model-free calibrated half-bleaching. · Muzzopappa F, Hummert J, Anfossi M, Tashev SA, Herten DP, Erdel F · Nat Commun 2022What this evidence cannot show
A methods paper. It constrains interpretation of condensate claims generally and does not itself test the SEC.
Evidence that argues against this (1)
- Clustered binding is a substantive alternative rather than a null; in a chromatin context it is arguably the more interesting model because it predicts sequence-specific targeting. Chong S et al. 2018
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
1,6-hexanediol rapidly immobilises and condenses chromatin in living human cells, an action distinct from melting a droplet, and the authors instruct readers to use this in reinterpreting chromatin-associated droplet results.
1,6-hexanediol rapidly immobilizes and condenses chromatin in living human cells. · Itoh Y, Iida S, Tamura S, Nagashima R, Shiraki K, Goto T, Hibino K, Ide S, Maeshima K · Life Sci Alliance 2021What this evidence cannot show
Live-cell imaging in human cell lines. It does not show that any specific published condensate is an artefact; it shows the reagent's action is not what the assay assumes.
Evidence that argues against this (1)
- Hexanediol insensitivity would still be informative in the other direction, so the reagent is not worthless - only unusable as positive evidence for a phase. Muzzopappa F et al. 2022
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
At physiological expression levels, low-complexity-domain interactions that control gene transcription form dynamic and selective hubs without detectable phase separation.
Imaging dynamic and selective low-complexity domain interactions that control gene transcription. · Chong S, Dugast-Darzacq C, Liu Z, Dong P, Dailey GM, Cattoglio C, Heckert A, Banala S, Lavis L, Darzacq X, Tjian R · Science 2018What this evidence cannot show
Single-molecule imaging in a defined set of transcription factors, not in the SEC. Absence of detectable phase separation at physiological expression does not prove absence in every system.
Evidence that argues against this (1)
- The SEC's multivalency is greater than that of a single transcription factor's activation domain, so the comparison is not exact and the SEC could behave differently. Guo C et al. 2020
proposed route by which SEC multivalency draws P-TEFb out of the soluble HEXIM1-bound pool
the assays used cannot separate a phase from clustered binding
the only cellular candidate for a directed extraction step