Oncofusion

start / graph / mechanism-q187-the-aff-degron-is-the-family-dosage-switch-and-three-paralogue-phenotypes-are-stabilisations

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q187-the-sec-family-and-yeats-selectivity

What the paralogues' disease biology says about AFF1: the AFF family shares one SIAH degron — and the index fusion DISCARDS it, because the whole degron lies in the AFF1 leader that the junction at residue 347 throws away

also called AFF family degron dosage control · PXAX-VXP Siah motif in the ALF family · the fusion loses the Af4 degron

All four of these scaffold proteins carry the same short disposal tag, and when that tag breaks the result is disease — a mouse with ataxia and two human syndromes. The leukaemia fusion protein still contains AFF1's tag, so the disposal route is intact in the oncoprotein and nobody has ever tried to use it.

The same thing, for a scientist

The dispatch question was whether the paralogues' own disease biology says anything about what AFF1 contributes to the fusion. **It does, and the axis is protein turnover rather than elongation.** Oliver and colleagues mapped the Siah-binding element of Af4 to a 23-residue stretch, **Af4 268-290**, containing a **PXAX-VXP** consensus motif; the *robotic* mouse's V280A sits exactly over that motif, reduces Siah-1a affinity roughly **15-fold**, almost abolishes Siah-1a-mediated degradation, raises transactivation about **2-fold** in a HeLa CAT reporter, and causes cerebellar ataxia with Purkinje loss. The homologous Fmr2 element is residues 277-299, and from 100% identity of the motif the authors **predict that all ALF-family members are Siah-1a substrates** [E1]. Human genetics hits the same switch twice more: KINSSHIP syndrome from de novo missense in, or deletion of, the nine-residue AFF3 **degron** that binds ubiquitin ligase [E2], and CHOPS syndrome from germline **gain-of-function** AFF4 missense [E3]. **The part that matters for this project, and which I got backwards before checking: the fusion does NOT escape by deletion.** The direct KMT2A::AFF1 joins to AFF1 exon 4, which begins at codon 54, so residues 268-290 are **retained** in the oncoprotein [E4]. `subtraction-aff1-control` reached this conclusion first and called the escape-by-deletion hypothesis a failure on coordinates. So the reading is inverted and improved: **the oncoprotein carries an intact, conserved E3-ligase recognition motif on the one axis where this protein family is demonstrably fragile, and no experiment has ever asked whether SIAH degrades KMT2A::AFF1.**

Go deeper
The motif, measured

From the full text of PMC522018, read 2026-09-24 1:

quantityvalue
Af4 Siah-binding peptideresidues 268-290 (23-mer, biotinylated, DELFIA)
Fmr2 homologous peptideresidues 277-299
consensus motifPXAX-VXP
robotic mutationV280A, sitting exactly over the motif
affinity loss≈15-fold decrease for mutant peptide against the Siah-1a substrate-binding domain
degradationSiah-1a-mediated degradation of mutant Af4 almost completely abolished
transactivation≈2-fold increase, HeLa CAT reporter; no significant difference in the yeast reporter
prediction100% identity of the PXAX-VXP motif across the family leads the authors to predict all ALF-family members are Siah-1a substrates

Two cautions the paper itself supplies and that any downstream use must carry. The affinity measurement is a peptide against an isolated substrate-binding domain, not two full-length proteins. And the transactivation effect is assay-dependent — 2-fold in HeLa, nothing in yeast — which is a small effect measured in one of two systems.

Three of four members have a gain-or-degron disease
memberlesion on the turnover axisphenotype
AFF1 (mouse robotic)V280A in the PXAX-VXP motif; Siah binding down ~15x, degradation abolishedcerebellar ataxia, Purkinje loss, cataracts 1
AFF2 / FMR2homologous motif at 277-299 binds Siah in the same way 1; the human disease is the opposite lesion, CCG-expansion silencingFRAXE intellectual disability
AFF3de novo missense in, or deletion of, the nine-residue ubiquitin-ligase-binding degronKINSSHIP syndrome 2
AFF4germline missense, titled gain-of-functionCHOPS syndrome 3

The pattern is not "the AFF family causes several diseases" — that is a list. It is that the recurrent pathogenic event in this family is failure of the turnover timer, and it recurs across a mouse dominant and two human syndromes found by three groups who were not looking for each other's result.

ana-q187-aff1-is-the-only-aff-scaffold-with-no-named-mendelian-syndrome carries the census behind the table: MedGen names a syndrome for AFF2, AFF3 and AFF4 and none for AFF1.

Reversed twice, and the plain reading was right: the fusion deletes the degron

This section asserted, in its first draft, that the fusion escapes SIAH by deleting the degron; then reversed to 'retained' on the strength of an exon-4-to-codon-54 mapping; then to 'undetermined'. Resolved this cycle from primary sequence: the first draft was right.

  • The junction is at AFF1 residue 347. QTYSNEVH occurs exactly once in UniProt P51825, at 347. Independently, ana-sima6-is-junction-spanning-and-it-pins-sems-breakpoint rebuilt the junction peptide VDFK|QTYSNEVHCVEEILK from the siMA6 siRNA and found that both KMT2A::AFF1 chimeras resume at QTYSNEVHCVEEILK — the AFF1 acceptor is constant, only the KMT2A donor varies. Two independent routes, both breakpoints, the same residue.
  • The degron is at human 282-304. The Siah coordinates 268-290 are mouse Af4. Mouse 268-290 is KSAAMQQKPTAYVRPMDGQDQAP; the human homologue is KSVAMQQKPTAYVRPMDGQDQAP at 282-304, 22/23 identical, offset +14. The PxAxVxP motif PTAYVRP is unique in human AFF1 at 290-296, and the residue corresponding to the robotic V280A is V294.
  • 304 < 347. The element ends 43 residues before the junction. The entire degron sits in the discarded AFF1 leader.

Why the corpus got the opposite answer. It placed the junction at 'residue ~53' from Ensembl exon-4 numbering and measured a 215-residue margin (268 - 53). The margin arithmetic was fine; the junction was wrong, and applying mouse coordinates to a human junction compounded it. subtraction-aff1-control explicitly stress-tested this — 'far larger than any plausible alignment error' — but stress-tested the alignment, which was never the weak term.

What it means. The subtraction hypothesis is back: KMT2A::AFF1 may be a stabilised protein by virtue of a deletion. And the forward and reciprocal products now differ in a clean way — the reciprocal AF4·MLL retains AFF1 1-346 and therefore the degron, yet Bursen and colleagues report it escapes proteasomal degradation anyway. So both products evade SIAH turnover, but only one of them does so by losing the motif.

The junction question is settled, and it settles it against this beat

fusion-kmt2a-aff1 and protein-aff1-domains were recorded as contradicting each other about the AFF1 junction. They do not, and fusion-kmt2a-aff1 was right.

  • bp-kmt2a-intron-11-aff1-e4 and bp-kmt2a-intron-9-10-aff1-e4 both set gene3_exon_first: 4, while domains_lost_3 reads "ALF / N-terminal homology domain (AFF1 1-346, upstream of the junction at residue 347) — CORRECTED 2026-09-20, this was listed as RETAINED".
  • That 347 is not an exon-numbering guess. The same entry records it as "verified by matching the junction peptide QTYSNEVH against the canonical P51825 FASTA", and that match is unique.
  • protein-aff1-domains had computed that under ENST00000307808 'exon 4' is codons 347-361 — the same residue, reached from exon structure rather than sequence. It listed "Which AFF1 codon does the fusion junction actually start at?" as the largest unresolved coordinate in its beat. It is now resolved: 347, and the three readings collapse to one physical exon boundary. The codon-54 reading is the only one refuted, and it is refuted by sequence.

This beat's earlier handling of that field was wrong twice: it first called the 347 a likely slip for exon 4's far end (353), then called the question undetermined. It was neither — 347 was the best-supported number in the corpus the whole time, carried by roughly twenty entries that use QTYSNEVH as the junction peptide.

Where this is weakest
  1. The AFF3 degron may not be the same element. The KINSSHIP abstract says "a nine amino acid sequence important for its binding to ubiquitin ligase" and does not name SIAH or give coordinates. Identifying it with the PXAX-VXP motif is this entry's inference from position and function. Oliver's prediction that all ALF members are Siah substrates makes the inference reasonable; it does not make it measured.
  2. The AFF3 knockout overlaps the human gain syndrome. Homozygous Aff3-null mice show skeletal, kidney, brain and neurological anomalies 2, overlapping KINSSHIP. The honest statement is bidirectional dosage sensitivity, not excess.
  3. CHOPS may not be a dosage lesion at all — its authors route the mechanism through altered AFF4/cohesin/RNAP2 binding and give no residue positions 3. If so, the pattern is two instances plus a mouse.
  4. Species and numbering. 268-290 and V280A are mouse Af4 coordinates from the peptide used in the assay. Human AFF1 (P51825, 1,210 aa) numbering was not independently checked here, and this graph has already recorded a separate 8-residue offset between UniProt P51825 and the Ensembl canonical AFF1 protein. Do not quote 268-290 as a human AFF1 range without re-deriving it.
Facts and figures
mechanism classprotein-dosage control by a shared E3-ligase degron, its failure as a family-level disease mechanism, and its RETENTION in the index oncoprotein
epigenetic readoutindirect. The axis sets the abundance of scaffolds that deliver P-TEFb and two YEATS acyl-lysine readers to chromatin, so a turnover defect is a dosage defect in epigenetic machinery rather than a mark change; no histone mark is measured in any source here.
shown for fusionNO for the direct fusion - no SIAH experiment has ever been run on KMT2A::AFF1. PARTIAL for the reciprocal: SIAH protection is reported for AF4.MLL (PMID 15221006).
measurements
transactivation_increase_…2.000transactivation_increase_of_the_robotic_af4_construct: 2aff_family_members_with_a…3.000aff_family_members_with_a_gain_or_degron_lesion: 3siah_binding_affinity_los…15.0siah_binding_affinity_loss_of_the_robotic_af4_peptide: 15
degron axis
{
  "ligases": [
    "Siah-1a (mouse, assayed)",
    "SIAH1",
    "SIAH2"
  ],
  "motif": "PXAX-VXP",
  "af4_element_residues": "268-290 (mouse Af4 peptide numbering, as synthesised in PMID 15459319)",
  "fmr2_element_residues": "277-299",
  "robotic_mutation": "V280A, sitting exactly over the motif",
  "affinity_loss_fold": 15,
  "members_assayed": [
    "AFF1 / Af4",
    "AFF2 / Fmr2"
  ],
  "members_predicted_not_assayed": [
    "AFF3",
    "AFF4"
  ],
  "retained_in_direct_fusion": false,
  "retained_in_direct_fusion_basis": "LOST. Junction fixed at AFF1 residue 347: QTYSNEVH occurs exactly once in UniProt P51825 (1,210 aa), and the corpus independently reconstructed the junction peptide VDFK|QTYSNEVHCVEEILK from the siMA6 siRNA, recording that both KMT2A::AFF1 chimeras resume at QTYSNEVHCVEEILK so the AFF1 acceptor is constant across breakpoints. Degron located in human numbering by homology: mouse Af4 268-290 (KSAAMQQKPTAYVRPMDGQDQAP) aligns 22/23 to human P51825 282-304 (KSVAMQQKPTAYVRPMDGQDQAP), offset +14; the PxAxVxP motif PTAYVRP is unique in human at 290-296 and the robotic-equivalent valine is V294. The element ends at 304, 43 residues N-terminal to the junction. Earlier drafts of this entry said retained (junction assumed at codon 54) and then undetermined; both are withdrawn.",
  "human_degron_coordinates": "AFF1 P51825 282-304; PxAxVxP motif PTAYVRP 290-296; robotic-equivalent V294",
  "murine_degron_coordinates": "Af4 O88573 268-290; PxAxVxP motif PTAYVRP 276-282; robotic V280A",
  "junction_residue": 347
}
lesion directions
{
  "AFF1_mouse_loss": "altered lymphoid development (PMID 10887138)",
  "AFF1_mouse_gain": "robotic: cerebellar ataxia, Purkinje loss, cataracts; V280A, Siah affinity down ~15x, degradation abolished, transactivation up ~2x (PMIDs 12629167, 15459319)",
  "AFF2_human_loss": "FRAXE intellectual disability (CCG expansion silencing)",
  "AFF3_human_degron": "KINSSHIP syndrome",
  "AFF3_mouse_loss": "skeletal, kidney, brain and neurological anomalies - OVERLAPS the human gain syndrome",
  "AFF4_human_gain": "CHOPS syndrome",
  "AFF1_somatic_in_KMT2A::AFF1": "degron LOST - the whole element (human 282-304) lies in the AFF1 leader discarded at the residue-347 junction. The reciprocal AF4-MLL retains it. Whether SIAH turnover matters to either product has never been tested."
}
marks named
[]
measurements — the numbers behind the chart
[
  {
    "quantity": "siah_binding_affinity_loss_of_the_robotic_af4_peptide",
    "subject": "Af4 268-290 wild-type vs V280A 23-mer peptide | Eu3+-labelled Siah-1a substrate-binding domain, DELFIA",
    "value": 15,
    "units": "fold decrease in affinity",
    "n": 1,
    "universe": "immobilised biotinylated 23-mer peptides, 36 pmol per well, 0-100 pmol labelled Siah-1a SBD, fluorescence at 613 nm; PMID 15459319 Fig 3",
    "null": 1,
    "note": "Stated in the paper as approximately 15-fold. A peptide against an isolated domain, not full-length proteins. Mouse Af4 numbering."
  },
  {
    "quantity": "transactivation_increase_of_the_robotic_af4_construct",
    "subject": "GAL4-BD Af4 N-terminal construct, robotic V280A vs wild-type",
    "value": 2,
    "units": "fold",
    "n": 1,
    "universe": "HeLa CAT reporter assay, PMID 15459319 Fig 7B; the parallel yeast ONPG reporter showed ~6-fold activation over empty vector for BOTH constructs and NO significant wild-type/mutant difference",
    "null": 1,
    "note": "ASSAY-DEPENDENT. This is the only evidence offered for the excess-activity reading of the robotic phenotype and it appears in one of two reporter systems."
  },
  {
    "quantity": "aff_family_members_with_a_gain_or_degron_lesion",
    "subject": "AFF1 (mouse robotic, degron), AFF2 (FRAXE - a LOSS lesion), AFF3 (KINSSHIP, degron), AFF4 (CHOPS, gain)",
    "value": 3,
    "units": "members of 4",
    "n": 4,
    "universe": "the four AFF/ALF scaffolds; counted if the primary report localises the variant to a degron, titles the lesion gain-of-function, or shows a dominant mutation abolishing E3-mediated degradation",
    "null": null,
    "note": "AFF2 is in the denominator, not the numerator - its human disease is gene silencing. If CHOPS proves to be a cohesin-interface rather than a dosage lesion, this becomes 2 of 4."
  }
]
How this could bite us back
  • The attractive reading of this entry is 'the fusion keeps a degron, so build a degrader through SIAH'. That inherits the family's dosage biology in the wrong direction as well as the right one: AFF1, AFF2 and AFF3 all carry the same motif and are all expressed in the developing brain, so any agent that enhances SIAH-mediated turnover of an ALF protein is a candidate to lower scaffolds whose loss causes FRAXE and whose knockouts have brain malformations. The germline genetics that makes this axis interesting is the same genetics that makes it dangerous.
  • The family pattern rests on one inferred joint - that the AFF3 KINSSHIP degron is the ALF PXAX-VXP motif - which neither source states. If that joint fails the entry is four genes with four diseases and no mechanism, and the joint is cheap to test by aligning the published variant positions.
  • This entry's first draft asserted the opposite conclusion - that the fusion escapes SIAH by deleting the degron - and it was wrong. It was caught by grepping the corpus, not by reasoning, and the corpus contained both the right answer and the wrong one in different entries. Any later reader should treat the retained-versus-lost question as live until `fusion-kmt2a-aff1` is reconciled, and should not assume this entry settled it.
  • The degron axis is orthogonal to the YEATS argument that motivated this beat and must not inherit its framing. Nothing here says YEATS inhibition perturbs AFF turnover; nobody has measured that, and the degron is a scaffold property with no counterpart in ENL or AF9.
  • This entry has now asserted retained, undetermined and lost in three successive drafts. A reader should weight the CURRENT claim by the method rather than the confidence: it rests on a unique peptide match and a 22/23 homology alignment, both reproducible in a few minutes, and it agrees with the flagship fusion entry that it earlier contradicted. But the underlying biology - whether SIAH turnover is load-bearing for either product - remains untested, so 'the fusion discards a degron' is a coordinate fact, not a demonstrated mechanism.
Still open
  1. Is KMT2A::AFF1 stabilised BECAUSE it lacks the degron? This is the subtraction hypothesis, which this corpus recorded as refuted on coordinates and which the corrected coordinates reinstate. The test is a turnover comparison between the fusion and an AFF1-leader-restored construct, and the reagents exist.
  2. Why is MLL-AF4 protein stable through the cell cycle in SEM cells if it retains an intact PXAX-VXP motif? Either the motif is not engaged in the fusion context, or SIAH activity is low in these cells, or turnover runs on a timescale the elutriation experiment could not see. Three distinguishable states, none tested.
  3. Where is the Siah motif in HUMAN AFF1 numbering? 268-290 and V280A are mouse Af4 peptide coordinates, and this corpus records an 8-residue offset between the Ensembl and UniProt AFF1 references. Any construct or degrader design needs the human range re-derived, and it has not been.
  4. Do AFF3 and AFF4 actually bind SIAH? Oliver and colleagues predict it from 100% motif identity; only Af4 and Fmr2 peptides were assayed. If the prediction fails, the 'shared family degron' is two members with a shared ligase rather than four.
  5. Is the AFF3 KINSSHIP degron the PXAX-VXP motif? The KINSSHIP abstract says only 'a nine amino acid sequence important for its binding to ubiquitin ligase'. Aligning the reported variant positions against the ALF motif would settle it from published data alone and would either strengthen or dissolve this entry's family pattern.
  6. Why does the reciprocal AF4-MLL escape proteasomal degradation while RETAINING the degron (Bursen et al.)? If retention does not predict turnover for the reciprocal, loss may not predict stabilisation for the forward fusion either, and the degron axis would be weaker than either the old or the new coordinate story implies.
  7. Does SIAH bind human AFF1 at 282-304? The element is assigned here by homology to murine Af4. No human SIAH-AFF1 binding map exists, and the corpus should stop quoting murine coordinates as though they were human.
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The Siah-binding element of Af4 is a 23-residue stretch at Af4 residues 268-290 containing the PXAX-VXP consensus motif, with the homologous Fmr2 element at residues 277-299; the robotic mouse mutation V280A sits exactly over the motif, reduces affinity of the mutant peptide for the Siah-1a substrate-binding domain approximately 15-fold, almost completely abolishes Siah-1a-mediated degradation of mutant Af4, and produces an approximately 2-fold increase in transactivation in a HeLa CAT reporter (with no significant difference in a yeast reporter); from 100% identity of the PXAX-VXP motif the authors predict that all members of the ALF family interact with and are degraded through Siah-1a-mediated ubiquitination.

    Mediation of Af4 protein function in the cerebellum by Siah proteins. · Oliver PL, Bitoun E, Clark J, Jones EL, Davies KE · Proceedings of the National Academy of Sciences of the United States of America 2004
    PMID 15459319DOI
    What this evidence cannot show

    Read from the PMC HTML full text of PMC522018 on 2026-09-24 (Europe PMC fullTextXML returns HTTP 500). Note the PMCID trap: Europe PMC's core record gives PMC522018, and a nearby id PMC522004 is a DIFFERENT paper that this beat fetched by mistake first - the title of the retrieved document must be read back before parsing it. Coordinates are MOUSE Af4 from the synthesised peptides; human AFF1 numbering was not re-derived. The affinity measurement is a 23-mer peptide against an isolated Siah-1a substrate-binding domain, not two full-length proteins. The transactivation effect is 2-fold, assay-dependent (present in HeLa CAT, absent in the yeast assay), and measured on GAL4-BD N-terminal constructs. The all-ALF-members statement is explicitly a PREDICTION from motif identity; only Af4 and Fmr2 peptides were assayed.

    Evidence that argues against this (1)
    • The paper's title frames Siah as MEDIATING Af4 function, not merely disposing of it, and the robotic allele changes a residue inside the motif rather than deleting the motif. A mutation that reduces binding to a partner and happens to stabilise the protein is consistent with the phenotype being caused by lost Siah signalling rather than by excess Af4 - a reading under which 'the degron is a dosage switch' misdescribes what the interaction does. The 2-fold transactivation increase is the only evidence offered for the excess-activity model and it is absent in one of the two reporter systems tried. Isaacs AM et al. 2003
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    KINSSHIP syndrome is caused by de novo missense variants in the degron of AFF3 — described by the authors as a nine amino acid sequence important for its binding to ubiquitin ligase — or by de novo deletions of that region, in eighteen individuals, partially overlapping AFF4-associated CHOPS syndrome; AFF3 overexpression in zebrafish caused body axis anomalies giving some support to a pathological effect of increased AFF3, while homozygous Aff3 knockout mice show skeletal anomalies, kidney defects, brain malformations and neurological anomalies.

    Variants in the degron of AFF3 are associated with intellectual disability, mesomelic dysplasia, horseshoe kidney, and epileptic encephalopathy. · Voisin N, Schnur RE, Douzgou S, Hiatt SM, Rustad CF, Brown NJ, Earl DL, Keren B, Levchenko O, Geuer S, Verheyen S, Johnson D, Zarate YA, Hančárová M, Amor DJ, Bebin EM, Blatterer J, Brusco A, Cappuccio G, Charrow J, Chatron N, Cooper GM, Courtin T, Dadali E, Delafontaine J, Del Giudice E, et al. · American Journal of Human Genetics 2021
    PMID 33961779DOI
    What this evidence cannot show

    Abstract only; the abstract gives no residue coordinates for the AFF3 degron and does not name SIAH as the ligase, so the identification of the AFF3 degron with the PXAX-VXP Siah motif characterised in E1 is this entry's INFERENCE from position and function, not a stated fact in either paper. That inference is the weakest joint in the entry and is flagged as such in the last section. The author list is truncated at 25 names as fetched.

    Evidence that argues against this (1)
    • The Aff3 knockout phenotype overlaps the human gain-of-function syndrome, which is the strongest internal objection to reading KINSSHIP as a pure stabilisation disease. If both directions of AFF3 dosage produce skeletal, renal and neurological anomalies, a degron variant's pathogenicity does not by itself demonstrate that stabilisation is the operative mechanism, and the family pattern has to be stated as bidirectional dosage sensitivity rather than as excess. Voisin N et al. 2021
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Germline missense mutations in AFF4, a core component of the super elongation complex, cause CHOPS syndrome in three unrelated probands and are reported as gain-of-function, with transcriptome and ChIP-seq analyses showing similar alterations of genome-wide AFF4, cohesin and RNAP2 binding in CHOPS and in Cornelia de Lange syndrome, and direct molecular interaction of SEC, cohesin and RNAP2 demonstrated.

    Germline gain-of-function mutations in AFF4 cause a developmental syndrome functionally linking the super elongation complex and cohesin. · Izumi K, Nakato R, Zhang Z, Edmondson AC, Noon S, Dulik MC, Rajagopalan R, Venditti CP, Gripp K, Samanich J, Zackai EH, Deardorff MA, Clark D, Allen JL, Dorsett D, Misulovin Z, Komata M, Bando M, Kaur M, Katou Y, Shirahige K, Krantz ID · Nature Genetics 2015
    PMID 25730767DOI
    What this evidence cannot show

    Three probands. 'Gain-of-function' is the paper's own framing in its title; the abstract states no residue positions and does not describe the variants as degron variants or as stabilising, so this entry counts CHOPS as a gain lesion on the dosage axis and explicitly does NOT count it as a degron lesion. Abstract only.

    Evidence that argues against this (1)
    • If the CHOPS mechanism is the cohesin interface rather than AFF4 abundance, this member drops out of the turnover pattern and the headline becomes two instances plus a mouse. The paper's own mechanistic emphasis - shared genome-wide binding changes with a cohesinopathy - points that way, and nothing in the abstract measures AFF4 protein level or half-life. Oliver PL et al. 2004
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    The AFF1 acceptor of KMT2A::AFF1 begins at AFF1 residue 347: the peptide QTYSNEVH occurs exactly once in UniProt P51825, and the junction peptide VDFK|QTYSNEVHCVEEILK reconstructed from the siMA6 siRNA shows both chimeras resuming at the same AFF1 acceptor. The Siah-binding element, whose published coordinates Af4 268-290 are murine, maps by 22/23 homology to human AFF1 282-304 with the PxAxVxP motif PTAYVRP unique at 290-296. The element therefore ends 43 residues N-terminal to the junction and is ABSENT from the oncoprotein.

    What this evidence cannot show

    Derived from primary sequence this cycle (UniProt P51825 and O88573, fetched 2026-09-24) combined with two corpus-internal results: the junction-peptide identity and the siMA6 reconstruction. It does NOT rest on exon numbering, which is what the earlier drafts got wrong. Residual dependencies: (i) that QTYSNEVH is genuinely the AFF1 acceptor rather than a coincidental corpus convention - supported by its use across roughly twenty entries and by the independent siMA6 reconstruction, but not re-derived here from a patient junction read; (ii) the human degron position is assigned by homology to the murine element, not measured - no one has mapped SIAH binding on human AFF1 directly, so 282-304 is an inference, albeit from a 22/23-identical alignment with an exactly conserved motif. Rarer breakpoints with a different AFF1 acceptor would need recomputing.

    Evidence that argues against this (1)
    • The field this entry judges to be wrong carries an explicit correction note - 'CORRECTED 2026-09-20, this was listed as RETAINED' - so somebody looked at exactly this question in an earlier cycle and reached the opposite conclusion deliberately. That is a reason to treat this as an unresolved dispute rather than a settled error, and to weight it accordingly: this beat has not seen whatever evidence drove that correction, and the 347 reading may rest on a source about the fusion junction that the exon-map argument does not address. Oncofusion Knowledge Graph 2026
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Bursen and colleagues report growth-transforming potential for AF4 wild-type and for the AF4.MLL fusion protein, and state that the underlying oncogenic mechanism involves the two E3 ubiquitin ligases SIAH1 and SIAH2, the N-terminal portion of AF4, and the protection of the AF4.MLL fusion protein against proteasomal degradation.

    Interaction of AF4 wild-type and AF4.MLL fusion protein with SIAH proteins: indication for t(4;11) pathobiology? · Bursen A, Moritz S, Gaussmann A, Moritz S, Dingermann T, Marschalek R · Oncogene 2004
    PMID 15221006DOI
    What this evidence cannot show

    THE SUBJECT IS THE RECIPROCAL PRODUCT AF4.MLL, not the direct fusion KMT2A::AFF1. The abstract does not state by what route AF4.MLL escapes degradation and this beat did not read the full text. Nothing in this item licenses any statement about whether the DIRECT fusion is a SIAH substrate - that question is untested in any cell and is this entry's central open question. 2004; transformation assays are cell-culture.

    Evidence that argues against this (1)
    • This paper's headline is that AF4 wild-type AND the reciprocal AF4.MLL are growth-transforming, a position on reciprocal-product oncogenicity that this corpus treats as contested and that most of the field does not adopt. Citing the SIAH mechanism while declining the transformation conclusion is a legitimate but partial reading and should be stated as such. It also complicates the simple picture here: if AF4 wild-type is itself transforming when overexpressed, then AFF1 dosage is oncogenic independently of any fusion, which would make the degron a tumour-suppressive element and the whole family story darker than the neurodevelopmental syndromes suggest. Oliver PL et al. 2004
Connected to