What the reciprocal product assembles: P-TEFb, an H3K4 methyltransferase, an H3K79 methyltransferase, an arginine methyltransferase and an acetyltransferase
also called AF4-MLL complex · AF4-MLL 16 constituents · reciprocal product P-TEFb
When the mirror-image fusion protein is purified out of cells it comes attached to at least sixteen other proteins, including the kinase that releases stalled RNA polymerase and three different enzymes that write marks on histones. That is a bigger machine than the normal AF4 complex.
The same thing, for a scientist
Affinity purification of AF4-MLL and of wild-type AF4 from cells found the AF4 complex to consist of 11 binding partners exhibiting P-TEFb-mediated activation of promoter-arrested RNA polymerase II with associated chromatin-modifying activities, while the AF4-MLL complex consisted of at least 16 constituents including P-TEFb kinase, H3K4(me3) and H3K79(me3) histone methyltransferases, a protein arginine N-methyltransferase and a histone acetyltransferase [E1]. The functional corollary was demonstrated on a specific gene: AF4-MLL strongly enhanced calcitriol-dependent transcript elongation of ALOX5 in a flavopiridol-sensitive manner, while the direct MLL-AF4 product had no effect on elongation at all [E2]. So the division of labour proposed by that laboratory is that the direct product acts at initiation and the reciprocal product at elongation. Two cautions apply hard here. The composition claim is 2011 affinity purification on protein this graph INFERS was over-expressed, of the same vintage and class as the fusion-composition claims this graph has already had to qualify. And crucially, **the H3K4me3 activity was not separated from the reciprocal product's own SET domain versus a co-purified SETD1A or KMT2B** - so the single most interesting possibility, that this is a catalytically active KMT2A fusion, is not established by this paper.
The elongation claim, on a named gene
ALOX5 transcription is controlled both at the promoter and by elongation within its 3' portion. The AF4 complex interacts directly with the vitamin D receptor and promotes calcitriol-dependent ALOX5 transcript elongation; AF4-MLL strongly enhanced that elongation and the enhancement was flavopiridol-sensitive, implicating P-TEFb; MLL-AF4 displayed no elongation effect 2. In a companion study the same group found MLL-AF4 enhanced ALOX5 promoter activity 47-fold, and that class I HDAC inhibition abrogated the dominant effect of MLL-AF4 by activating wild-type MLL 3. Together these are a promoter/elongation split between the two products on one gene, which is a cleaner mechanistic statement than anything genome-wide in this literature.
The RUNX1 connection is the only independent structural claim about this complex
An independent laboratory reported that RUNX1 is directly activated by MLL-AF4 and that the RUNX1 protein interacts with the reciprocal AF4-MLL product, proposing a mechanism in which the two fusion proteins cooperate by activating a target gene and then modulating the function of its product 4. That is the only report from outside the Frankfurt group placing a named protein on the reciprocal complex. Its own limitation is the reagent problem: the AF4-MLL species was assigned by apparent molecular weight in an anti-MLL-C immunoprecipitation, and that antibody class cannot distinguish reciprocal-derived MLL-C from wild-type MLL-C.
| complex size af4 mll | 16 |
|---|---|
| complex size af4 wt | 11 |
| named interactor from independent lab | RUNX1 |
| worked target gene | ALOX5 |
| unresolved | whether the H3K4me3 activity is the reciprocal product's own SET domain or a co-purified SET1-family enzyme |
| method class | affinity purification, 2011; never checked by proximity labelling |
activities reported
[ "P-TEFb kinase", "H3K4me3 methyltransferase", "H3K79me3 methyltransferase", "protein arginine N-methyltransferase", "histone acetyltransferase" ]
- If the reciprocal product carries active H3K4 methyltransferase activity, then a t(4;11) cell contains a KMT2A fusion that a menin inhibitor cannot reach and that still writes the activating mark - and nobody has tested whether that sustains target genes under menin blockade.
- The flavopiridol sensitivity of the reciprocal product's elongation activity suggests a CDK9 combination, but CDK9 inhibition is broadly cytotoxic and the selectivity argument would have to come from elsewhere.
- Every composition claim here is 2011 affinity purification on protein this graph INFERS was over-expressed from one laboratory; an orthogonal method could change the picture entirely.
- Is the AF4-MLL complex's H3K4me3 activity the fusion's own SET domain? A catalytically dead SET mutant of the reciprocal product would answer it and has never been made.
- Which 16 proteins? The composition is described by activity class rather than enumerated, so no subunit-level entry can be written from it.
- Has any orthogonal method - proximity labelling, IP-MS with a junction-specific reagent - ever been applied to the reciprocal product? No KMT2A-complex protein has ever been a proximity bait at all.
- Does the promoter/elongation division of labour hold on any gene other than ALOX5?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Affinity purification found the AF4 complex to consist of 11 binding partners with P-TEFb-mediated activation of promoter-arrested RNA polymerase II, and the AF4-MLL complex to consist of at least 16 constituents including P-TEFb kinase, H3K4(me3) and H3K79(me3) histone methyltransferases, a protein arginine N-methyltransferase and a histone acetyltransferase.
The leukemogenic AF4-MLL fusion protein causes P-TEFb kinase activation and altered epigenetic signatures. · Benedikt A, Baltruschat S, Scholz B, et al. · Leukemia 2011What this evidence cannot show
2011 affinity purification on protein this graph INFERS was ectopically expressed; the constituents are described by activity class in the abstract rather than enumerated by gene symbol, and the H3K4me3 activity is not attributed to the fusion's own SET domain as opposed to a co-purified SET1-family enzyme. Same laboratory as the model in which the reciprocal product is leukaemogenic. **Correction 2026-09-20: the expression system is an inference, not a read fact.** PMID 21030982's abstract does not state it, and the full text fails all four routes - no PMCID, no PMC record, and the publisher page serves abstract, paywall and references only. The murine HSPC model makes ectopic expression a reasonable inference, and it should be labelled as one.
Evidence that argues against this (2)
- Affinity purification and proximity labelling recover largely non-overlapping partner sets by design, and no KMT2A-complex protein has ever been used as a proximity bait, so a 2011 affinity-purification composition is one-directional evidence that has never been checked by an orthogonal method. Wilkinson AC et al. 2013
- An unread methods section is the commonest way a caveat becomes a fact in this graph. Nothing about the expression system in this purification has been verified, and the entry previously asserted it in two places. Leukemia 2011
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The AF4 complex interacts directly with the vitamin D receptor and promotes calcitriol-dependent ALOX5 transcript elongation; activation of elongation was strongly enhanced by AF4-MLL and was flavopiridol-sensitive, whereas MLL-AF4 displayed no effect on transcriptional elongation, and class I HDAC inhibitors blocked the ectopic effects of AF4-MLL.
AF4 and AF4-MLL mediate transcriptional elongation of 5-lipoxygenase mRNA by 1, 25-dihydroxyvitamin D3. · Ahmad K, Scholz B, Capelo R, et al. · Oncotarget 2015What this evidence cannot show
A reporter and transcript-elongation system on one gene in a non-leukaemic context; flavopiridol is a broad CDK inhibitor, so P-TEFb attribution is pharmacological rather than genetic; the HDAC inhibitor result is a proposed therapeutic inference from a single-gene assay.
Evidence that argues against this (1)
- The companion study from the same group attributes the HDAC-inhibitor effect to reactivation of wild-type MLL overriding MLL-AF4, not to an action on AF4-MLL, so the therapeutic inference for t(4;11) is drawn from two different mechanisms in two papers. Ahmad K et al. 2014
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
MLL-AF4 enhanced ALOX5 promoter activity 47-fold, further stimulated by co-transfected vitamin D receptor with RXR or by SMAD3/SMAD4; combined inhibition of HDAC1-3 induced the ALOX5 promoter in an MLL-dependent manner and inhibited the constitutive activation caused by MLL-AF4 by relieving inhibitory functions of wild-type MLL, while MLL knockdown increased MLL-AF4-mediated effects.
Inhibition of class I HDACs abrogates the dominant effect of MLL-AF4 by activation of wild-type MLL. · Ahmad K, Katryniok C, Scholz B, et al. · Oncogenesis 2014What this evidence cannot show
Promoter-reporter transfection experiments; the dominant-effect framing concerns the direct product, and the entry cites it here for the promoter half of the promoter/elongation division of labour, not for a claim about the reciprocal product.
Evidence that argues against this (1)
- A single-gene promoter-reporter result does not establish a genome-wide division of labour; the same laboratory's genome-wide experiment in HEK293T attributes broad chromatin opening to the reciprocal product rather than a promoter/elongation split. Wilhelm A et al. 2021
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
RUNX1 is directly activated by MLL-AF4 and the RUNX1 protein interacts with the product of the reciprocal AF4-MLL translocation, supporting a mechanism in which the two fusion proteins cooperate by activating a target gene and then modulating the function of its downstream product.
RUNX1 is a key target in t(4;11) leukemias that contributes to gene activation through an AF4-MLL complex interaction. · Wilkinson AC, Ballabio E, Geng H, et al. · Cell Rep 2013What this evidence cannot show
The reciprocal species was assigned in anti-MLL-C and anti-AF4-N immunoprecipitations by apparent molecular weight (~194 kDa); that reagent class cannot distinguish reciprocal-derived MLL-C from wild-type MLL-C, so the interaction could in principle be with wild-type MLL-C in the same complex.
Evidence that argues against this (1)
- A fusion-versus-wild-type discrimination by antibody is achievable for a direct fusion - an MLL1 C-terminal antibody does not detect MLL-AF9 - but not for a reciprocal fusion that retains the identical C-terminus, so this attribution is structurally unresolvable with current reagents. Xu J et al. 2016
the 11-and-16-component inventory rests on a paper unreadable by all four routes, and describes the reciprocal product rather than KMT2A::AFF1
the reciprocal product is one of the C-only species that the MLL-C track cannot separate from wild-type
records the four-route failure behind that entry's only composition source, and bounds what may be asserted from it
a mechanistic explanation for why AF4-MLL drives elongation and MLL-AF4 does not
PARTNER-IMPORT GUARD: the writer-rich inventory (H3K4me3 and H3K79me3 methyltransferases, a PRMT, a HAT) belongs to AF4-MLL, the RECIPROCAL derivative, not to KMT2A::AFF1. Recorded on the AFF1 entry so the epigenetic column is not filled from a naming collision.
The multiprotein assembly the reciprocal oncoprotein nucleates.
P-TEFb kinase reported as a constituent; the elongation effect on ALOX5 is flavopiridol-sensitive.
An H3K79(me3) methyltransferase activity is reported among the 16 constituents.
RUNX1 protein interacts with the reciprocal product, per an independent laboratory; the species assignment rests on apparent molecular weight.
Reports H3K4me3 methyltransferase activity in the reciprocal complex, without separating the fusion's own SET domain from a co-purified enzyme.