Taspase1 is what turns the reciprocal product into an oncoprotein - and it is a conditional oncogene because of it
also called Taspase1 and AF4-MLL · TASP1 cleavage of the der(4) product · conditional oncoprotein
The mirror-image fusion protein is made in one long piece that the cell quickly destroys. An enzyme called Taspase1 cuts it in two, and the two halves clip back together into a package that survives for a very long time. Without that cut, there is no oncoprotein.
The same thing, for a scientist
AFF1::KMT2A retains both KMT2A Taspase1 consensus sites (UniProt Q03164 cleavage sites at 2666/2667 and 2718/2719) [E4]. Cleavage is not incidental: the uncleaved 328 kDa AF4-MLL protein is a proteasome substrate, while the cleaved form assembles into a stable multiprotein complex with a very long half-life, and a dominant-negative Taspase1 causes the leukaemogenic oncoprotein to disappear [E1]. Taspase1 is therefore described by that group as a conditional oncoprotein in t(4;11). The catalytic route is the druggable one: catalytically dead Taspase1 variants over-expressed nine-fold do not inhibit the wild-type enzyme in trans, because inactive subunits fail to assemble stable heterocomplexes with it [E2], so a dimerisation-blocking strategy is disfavoured and an active-site or allosteric strategy is favoured. One allosteric inhibitor exists - closantel sodium, IC50 1.6 microM in an HTRF reporter assay - and its authors position it as a tool for target validation, not a drug [E3]. The corollary this graph cares about is the assay one: because the reciprocal protein is cleaved at exactly the wild-type sites, the MLL-C fragment it releases is sequence-identical to the wild-type MLL-C fragment.
Why cleavage is the on-switch rather than a processing detail
Two states of the same protein behave oppositely. Uncleaved AF4-MLL (328 kDa) is degraded by the proteasome; cleaved AF4-MLL forms a stable high-molecular-weight complex whose half-life is very long 1. So Taspase1 activity, not transcription of the der(4) allele, is the step that determines whether a reciprocal-positive cell actually carries a reciprocal oncoprotein. This matters for the clinical literature, where every reciprocal-status measurement is RT-PCR: a transcript-positive patient whose blasts have low Taspase1 activity is, on this model, protein-negative, and nobody has measured that.
Why the obvious inhibitor design does not work
The intuitive move against a dimerising protease is a dominant-negative subunit. It was tried and it failed: Taspase1(T234V) and Taspase1(D233A) at nine-fold over-expression inhibited neither cis- nor trans-cleavage, and separately expressed alpha- and beta-subunits did not assemble into active enzyme; the inactive protein simply does not form stable heterocomplexes with the wild-type enzyme 2. The authors' own conclusion is to target catalytic activity instead. Note that this sits in tension with 1 from the same field, which reports that a designed dominant-negative Taspase1 does cause AF4-MLL to disappear - the two papers use different constructs and different readouts, and this graph carries both rather than choosing.
What is actually available as chemical matter
Closantel sodium, a veterinary anthelmintic, inhibits Taspase1 non-covalently with IC50 1.6 microM by interfering with dimerisation and/or the proenzyme's intrinsic serine protease function 3. That is a probe, not a lead: micromolar, repurposed, and aimed at a protease whose wild-type substrates include KMT2A itself, KMT2B/MLL4 and TFIIA 1. Inhibiting Taspase1 in a t(4;11) cell would stop processing of the wild-type KMT2A allele at the same time.
| kind | post-translational activation |
|---|---|
| protease | TASP1 (Taspase1) |
sites on kmt2a
[
[
2666,
2667
],
[
2718,
2719
]
]chemical matter
{
"compound": "closantel sodium",
"ic50_uM": 1.6,
"assay": "HTRF reporter",
"status": "tool compound"
}products
[ "AF4-MLL-N (AFF1 N-terminus + KMT2A to 2666)", "MLL-C (KMT2A 2719-3969)" ]
wild type substrates also cleaved
[ "KMT2A", "KMT2B/MLL4", "TFIIA" ]
- Taspase1 cleaves wild-type KMT2A and KMT2B and TFIIA, so inhibiting it removes normal MLL processing at the same time as fusion processing - an essential-function toxicity built into the mechanism.
- The only reported inhibitor is a micromolar repurposed anthelmintic with no published anti-leukaemic cell data in t(4;11).
- If a Taspase1 inhibitor worked, it would leave the direct MLL-AF4 product completely untouched, so it could only ever be a combination agent with a menin inhibitor - and that combination has never been tested.
- The two published dominant-negative results contradict each other, so the mechanistic basis for any Taspase1 programme in this disease is not settled.
- Is Taspase1 activity in primary t(4;11) blasts high enough to cleave the reciprocal product, and does it vary between reciprocal-transcript-positive patients who do well and those who do not?
- Do the two papers disagreeing about dominant-negative Taspase1 disagree about the enzyme or about their constructs? Nobody has run both designs side by side.
- Would Taspase1 inhibition in a t(4;11) cell be tolerable, given that wild-type KMT2A, KMT2B and TFIIA are its physiological substrates?
- Does a Taspase1 inhibitor differentially affect reciprocal-positive versus reciprocal-negative t(4;11) lines? That is the experiment that would make Taspase1 a reciprocal-selective handle rather than a general one.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Taspase1 cleaves MLL, MLL4 and TFIIA at their consensus sites and AF4-MLL in the leukaemogenic case; the uncleaved 328 kDa AF4-MLL protein is subject to proteasomal degradation while the cleaved form assembles a stable oncogenic multiprotein complex with a very long half-life, and a designed dominant-negative Taspase1 causes the leukaemogenic oncoprotein to disappear.
Unraveling the Activation Mechanism of Taspase1 which Controls the Oncogenic AF4-MLL Fusion Protein. · Sabiani S, Geppert T, Engelbrecht C, et al. · EBioMedicine 2015What this evidence cannot show
Site-directed mutagenesis and a BFP-CSTasp1-GFP FRET biosensor in transfected cells; the half-life and stability claims are on ectopically expressed protein, not on an endogenous fusion in a patient-derived line, and the group reporting this is the group that proposed the reciprocal product's oncogenic role.
Evidence that argues against this (1)
- A directly contradictory result on the same protein class: nine-fold over-expression of catalytically dead Taspase1(T234V) or Taspase1(D233A) inhibited neither cis- nor trans-cleavage in AF4-MLL model systems, because the inactive protein fails to form stable heterocomplexes with the wild-type enzyme. Bier C et al. 2012
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Nine-fold over-expression of catalytically dead Taspase1(T234V) or attenuated Taspase1(D233A) failed to inhibit Taspase1 cis- or trans-cleavage of AF4-MLL in living cells, separately expressed alpha- and beta-subunits did not assemble into active protease, and inefficient heterodimerisation is the reason; the authors favour targeting catalytic activity over blocking dimer formation.
Overexpression of the catalytically impaired Taspase1 T234V or Taspase1 D233A variants does not have a dominant negative effect in T(4;11) leukemia cells. · Bier C, Hecht R, Kunst L, et al. · PLoS One 2012What this evidence cannot show
Cell-based assay in model systems rather than in primary t(4;11) blasts; a negative result about a particular pair of mutants, which does not exclude other dominant-negative designs.
Evidence that argues against this (1)
- The opposite result is published from the same field: a designed dominant-negative Taspase1 co-expressed with AF4-MLL caused the oncoprotein to disappear, with uncleaved AF4-MLL degraded by the proteasome. Sabiani S et al. 2015
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Closantel sodium inhibits human Taspase1 non-covalently with IC50 1.6 microM, interfering with the dimerisation step and/or the intrinsic serine protease function of the proenzyme, and is presented as a means of further validating Taspase1 as a cancer target.
Closantel is an allosteric inhibitor of human Taspase1. · Luciano V, Proschak E, Langer JD, et al. · iScience 2021What this evidence cannot show
A reporter-assay hit at micromolar potency in a repurposed veterinary anthelmintic; no cellular anti-leukaemic data in a t(4;11) system, no selectivity panel over other proteases reported in the abstract, and the authors frame it as a validation tool.
Evidence that argues against this (1)
- The same paper's mechanism - interference with dimerisation - is the strategy an independent study concluded does not work in cells, because inactive Taspase1 does not form stable heterocomplexes with the active enzyme. Bier C et al. 2012
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt Q03164 annotates two Taspase1 cleavage sites on KMT2A, at residues 2666/2667 and 2718/2719, both of which lie C-terminal to the t(4;11) breakpoint cluster region and are therefore retained in the reciprocal product.
UniProt Q03164 (KMT2A_HUMAN), feature table · UniProt Consortium · UniProtKB 2026What this evidence cannot show
Fetched from the UniProt REST record 2026-09-19, primary gene name read back as KMT2A, length 3969. UniProt annotates the sites on the wild-type chain; the inference that they are retained in AFF1::KMT2A follows from the exon composition reported elsewhere and is not itself a UniProt statement.
Evidence that argues against this (1)
- Retention of a site is not proof of cleavage in situ: the reciprocal product has never been shown to be cleaved in a patient blast, and the one immunoprecipitation assigning a cleaved ~194 kDa AF4-MLL species in t(4;11) lines did so on apparent molecular weight rather than by an epitope that distinguishes it from wild-type MLL-C. Wilkinson AC et al. 2013
Cleavage at KMT2A 2666/2667 and 2718/2719 converts a proteasome substrate into a long-lived complex; without it there is no reciprocal oncoprotein.
- regulates →KMT2A / MLL1 — the gene that breaks
The same protease cleaves wild-type KMT2A at the same two sites, which is both the toxicity risk and the reason the released MLL-C fragments are indistinguishable.
Identical cleavage sites are the mechanical reason the released MLL-C fragments cannot be told apart.
Cleavage creates the two fragments that the FYRN/FYRC interface then re-associates.
- ← related toSIAH1/SIAH2: the disposal route that destroys wild-type AF4 and that the reciprocal fusion escapes
A second, unreconciled account of the same protein's stability.