Oncofusion

start / graph / mechanism-the-ptefb-binding-module-of-aff1-travels-on-the-reciprocal-product

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q35-ptefb-reservoir

The two t(4;11) products split AFF1 exactly at the P-TEFb-binding module, and the elongation half goes to the reciprocal

also called AF4N on the reciprocal · module partition at the t(4;11) junction

The translocation cuts AFF1 in two. The front piece, which is the part that grips the elongation enzyme, ends up in the reciprocal product; the back piece, which grips the other scaffold proteins, ends up in the main fusion. That division explains an old observation nobody had an explanation for: only the reciprocal product actually drives elongation.

The same thing, for a scientist

The functional module map of AFF1 and the position of the t(4;11) junction coincide almost exactly, and the coincidence has consequences this graph has not previously drawn. The AFF1 N-terminal homology domain plus ALF domain - operationally AF4N, aa 1-360 - are necessary and sufficient to recruit P-TEFb [E1][E2]; the cyclin T1-binding sequence within them is aa 1-308 with a minimal element at 38-72 [E1]. The junction sits at **AFF1 residue 347** [E3][E4]. Therefore the direct product KMT2A::AFF1 carries AFF1's C-terminal half - the pSER region and the CHD that binds AF9/ENL and mediates AFF-family heterodimerisation - and the reciprocal product AFF1::KMT2A carries the P-TEFb-binding module. This predicts an asymmetry that was already observed and unexplained: affinity purification found the AF4-MLL complex to contain P-TEFb among at least 16 constituents, and AF4-MLL strongly enhanced calcitriol-dependent ALOX5 transcript elongation in a flavopiridol-sensitive manner **while the direct MLL-AF4 product had no effect on elongation at all** [E5]. The module partition is a mechanistic explanation for that result. It also means the reciprocal product, not the direct one, is the t(4;11) protein that could in principle interact with the 7SK reservoir.

Go deeper
The module map against the junction
AFF1 moduleresiduesfunctionwhich t(4;11) product
NHD + ALF (AF4N)~1-360recruits P-TEFb and other factorsreciprocal AFF1::KMT2A
cyclin T1-binding sequence1-308, minimal 38-72docks on cyclin T1; entry into the 7SK snRNPreciprocal
Siah-binding degron268-290 (as recorded elsewhere in this graph)proteasomal turnoverreciprocal
junction347--
pSER transactivation regionC-terminal to the junctiontransactivationdirect KMT2A::AFF1
CHDC-terminalAF9/ENL binding; AFF-family heterodimerisationdirect

The clean reading is that t(4;11) does not merely relocate AFF1's C-terminus onto KMT2A; it separates AFF1's elongation-factor-binding half from its scaffold-binding half and puts them on different proteins. In a reciprocal-positive patient both halves are present, on two molecules that can in principle be bridged - the direct product's CHD heterodimerises with AFF-family partners. In a reciprocal-negative patient only the scaffold half is present.

What this predicts, and what would kill it

Predicts:

  • Reciprocal-positive t(4;11) blasts have more P-TEFb-driven elongation at fusion target genes than reciprocal-negative ones, other things equal.
  • CDK9-inhibitor sensitivity differs by reciprocal status.
  • A co-IP of the reciprocal product pulls down HEXIM1 and LARP7; a co-IP of the direct product does not.

Would kill it:

  • A co-IP showing the direct product bound to P-TEFb without a bridging AFF4 or wild-type AFF1.
  • No difference in Pol II CTD Ser2 at HOXA9 between reciprocal-positive and reciprocal-negative blasts.

This graph already records that no menin-inhibitor study has stratified by reciprocal status and that reciprocal expression predicts better survival in t(4;11) infants. If the reciprocal product carries the elongation module, better survival in reciprocal-positive patients is at first sight the wrong direction, and that tension is worth stating rather than smoothing over: either the elongation module is not rate-limiting in patients, or the reciprocal's effect on outcome runs through something else entirely. This entry does not resolve it.

Facts and figures
mechanism classmodule partition at a translocation junction
junction residue aff1347
explainsthe reported asymmetry in which AF4-MLL enhances flavopiridol-sensitive elongation at ALOX5 while MLL-AF4 does not
unresolved tensionthis graph records that reciprocal expression predicts BETTER survival in t(4;11) infants, which is the opposite direction from 'the reciprocal carries the elongation module'
direct product carries
[
  "AFF1 347-1210",
  "pSER transactivation region",
  "CHD (AF9/ENL binding, AFF-family heterodimerisation)"
]
reciprocal product carries
[
  "AFF1 1-346",
  "NHD + ALF (most of AF4N)",
  "cyclin T1-binding sequence 1-308",
  "Siah-binding degron - murine Af4 268-290, human AFF1 282-304 (22/23 homology, motif PTAYVRP 290-296); inside AFF1 1-346 on either numbering, so the reciprocal carries it"
]
predictions
[
  "reciprocal-positive blasts show more Pol II CTD Ser2 at fusion targets",
  "CDK9-inhibitor sensitivity differs by reciprocal status",
  "the reciprocal product co-IPs HEXIM1 and LARP7 and the direct product does not"
]
related entries
[
  "protein-aff1-kmt2a-reciprocal-product",
  "complex-af4-mll-assembles-a-sixteen-component-elongation-machine",
  "concept-no-menin-inhibitor-study-has-stratified-by-reciprocal-status",
  "epi-reciprocal-expression-predicts-better-survival-in-t411-infants",
  "mechanism-kmt2a-aff1-cannot-be-sequestered-in-the-7sk-snrnp"
]
How this could bite us back
  • If the elongation-competent module is on the reciprocal product, then a CDK9-directed strategy is aimed at the allele that half of patients do not retain, and reciprocal status would have to become a stratification variable that no trial currently collects.
Still open
  1. Does a co-IP of the reciprocal AFF1::KMT2A product pull down HEXIM1, LARP7 and 7SK? Nobody has asked.
  2. Does CDK9-inhibitor sensitivity in t(4;11) cells track reciprocal status? Enitociclib was tested in four KMT2A-rearranged lines whose reciprocal status is not reported.
  3. If the reciprocal carries the elongation module, why does reciprocal expression associate with better survival? Either the module is not rate-limiting in patients, or the survival association runs through something else.
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    AFF1's incorporation into the 7SK snRNP and its binding to P-TEFb require the N-terminal cyclin T1-binding sequence: deletion of residues 1-308 abolishes binding to P-TEFb, HEXIM1 and LARP7 while preserving ELL2 and ENL contacts, and AFF1 1-77 or 38-72 alone suffices.

    AFF1 is a ubiquitous P-TEFb partner to enable Tat extraction of P-TEFb from 7SK snRNP and formation of SECs for HIV transactivation. · Lu H, Li Z, Xue Y, Schulze-Gahmen U, Johnson JR, Krogan NJ, Alber T, Zhou Q · Proc Natl Acad Sci U S A 2014
    PMID 24367103DOI
    What this evidence cannot show

    HeLa co-IP with overexpressed deletion constructs, in an HIV transactivation study.

    Evidence that argues against this (1)
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The AF4-family N-terminal homology domain and ALF domain - operationally AF4N, AFF1 aa 1-360 - are necessary and sufficient to recruit P-TEFb and other factors, while the C-terminal portion binds AF9 or ENL, carries a nuclear localisation sequence and confers heterodimerisation.

    DDX6 transfers P-TEFb kinase to the AF4/AF4N (AFF1) super elongation complex. · Muck F, Bracharz S, Marschalek R · Am J Blood Res 2016
    PMID 27679741
    What this evidence cannot show

    Stated as established background in a single-laboratory paper and supported there by overexpression work in HEK 293T. Independently consistent with the deletion mapping in E1, which is the reason it is carried at moderate rather than weak.

    Evidence that argues against this (1)
    • The same paper's headline mechanism (DDX6-mediated transfer) is unreplicated, so its framing of the module map should be checked against E1 rather than relied on alone - which is what this entry does. Lu H et al. 2014
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    All three recurrent KMT2A breakpoint architectures splice to the same AFF1 acceptor and produce the AFF1 head peptide QTYSNEVH.

    What this evidence cannot show

    An internal computation from RefSeq transcripts, covering the three UniProt-annotated recurrent breakpoints and not rare variant junctions.

    Evidence that argues against this (1)
    • This graph's fusion entry records one case in 1,262 with a genomic break upstream of AFF1 exon 1, so the acceptor is essentially but not absolutely invariant. Oncofusion Knowledge Graph et al. 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    QTYSNEVH occurs once in UniProt P51825 beginning at residue 347, so the direct fusion carries AFF1 347-1210 and the reciprocal product carries AFF1 1-346.

    What this evidence cannot show

    A sequence string match run in this session against the FASTA fetched the same day; isoform choice shifts numbering by a few residues, far less than the 38-residue margin between 308 and 347.

    Evidence that argues against this (1)
    • AF4N as defined by Mueck et al. is 1-360, so residues 347-360 of that module do fall on the direct product's side; the functionally mapped cyclin T1-binding sequence does not. Muck F et al. 2016
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    Affinity purification found the AF4-MLL complex to comprise at least 16 constituents including P-TEFb kinase, and AF4-MLL strongly enhanced calcitriol-dependent ALOX5 transcript elongation in a flavopiridol-sensitive manner while the direct MLL-AF4 product had no effect on elongation.

    What this evidence cannot show

    This cites a sibling entry in this knowledge graph, which itself records the underlying work as 2011-vintage affinity purification on probably-overexpressed protein, never checked by proximity labelling. No identifier has been copied from that entry; only the substance of its claim is used, and the primary sources it lists were not re-fetched by this beat.

    Evidence that argues against this (1)
    • This graph separately records that the reciprocal product is dispensable for maintenance, so whatever elongation activity AF4-MLL has in a reporter assay is not obviously required by an established leukaemia. Oncofusion Knowledge Graph et al. 2026
Connected to