SIAH1 — the E3 ligase that degrades AFF1, and that the reciprocal AF4::KMT2A fusion escapes

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

• SIAH1 and SIAH2 are broadly expressed E3 ligases with many substrates including beta-catenin regulators; any attempt to modulate them systemically would have wide consequences.
• If the reciprocal AFF1::KMT2A product is not leukemogenic in human cells, as the human CD34+ data suggest, then a strategy built on this interaction targets a protein that may not be driving the disease.

1. Where exactly in AFF1 is the SIAH degron? Without residue boundaries it is impossible to check formally whether any KMT2A::AFF1 breakpoint variant retains it, and the exclusion rests on the phrase 'N-terminal portion'.
2. Has any group outside Frankfurt replicated the SIAH1/SIAH2-AF4 interaction? Nothing was retrieved, and a single-group interaction that has stood for twenty years without replication should be flagged as such.
3. Is wild-type AFF1 protein abundance in t(4;11) blasts actually elevated, as the stabilisation model predicts for the fusion but not for the wild-type allele?
4. Could a molecular glue be built to re-impose SIAH-mediated degradation on KMT2A::AFF1 by grafting recognition onto a surface the der(11) product does retain? This is the degrader question the entry points at and nobody appears to have asked it.

1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   The E3 ubiquitin ligases SIAH1 and SIAH2 interact with the N-terminal portion of AF4/AFF1 and target wild-type AF4 for proteasomal degradation, while the AF4.MLL fusion protein is protected from that degradation; both AF4 wild type and the AF4.MLL fusion protein show growth-transforming potential.

   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

   Abstract read in full; not full text. The title itself ends in a question mark, which is the authors' own signal about how far the pathobiological claim reaches. The interaction is mapped only to 'the N-terminal portion of AF4' with no residue boundaries, no structure, and no reported affinity. No independent replication by a second group was retrieved. Everything downstream in this entry that depends on the contact being N-terminal depends on that one phrase.

   ▸Evidence that argues against this (1)

   • The functional importance of the reciprocal AF4::MLL product, which is what makes the SIAH interaction interesting, does not reproduce in human cells. AF4-MLL expressed in human cord-blood CD34+ HSPCs augmented multilineage engraftment transiently but conferred no serial self-renewal and initiated no leukemia within 36 weeks, in direct contrast to the murine result. If the reciprocal product is not leukemogenic in human cells, SIAH-mediated stabilisation of it is a mouse phenomenon. Prieto C et al. 2017
2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   Murine Lin-/Sca1+ progenitors transduced with AF4.MLL alone, or with both fusions, developed pro-B ALL, B/T biphenotypic acute leukemia or mixed lineage leukemia, whereas MLL.AF4-transduced or mock-transduced cells produced no disease over 13 months of observation.

   The AF4.MLL fusion protein is capable of inducing ALL in mice without requirement of MLL.AF4. · Bursen A, Schwabe K, Rüster B, Henschler R, Ruthardt M, Dingermann T, Marschalek R · Blood 2010
   PMID 20194896DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. Retroviral overexpression in murine progenitors, which is a high-expression artificial system; the negative result for MLL.AF4 in this assay is widely understood to reflect the well-documented difficulty of modelling MLL-AF4 in mice rather than an absence of leukemogenicity, since MLL::AFF1 unambiguously causes leukemia in humans. Same group as E1.

   ▸Evidence that argues against this (1)

   • The same group's later work in human cells did not reproduce the leukemogenicity of AF4-MLL, and the authors attribute the difference to species-specific determinants. Prieto C et al. 2017
3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   AF4-MLL expressed in human cord-blood CD34+ HSPCs enhanced long-term multilineage reconstitution in primary immunodeficient recipients but conferred no self-renewal on serial transplantation and initiated no leukemia within 36 weeks.

   The AF4-MLL fusion transiently augments multilineage hematopoietic engraftment but is not sufficient to initiate leukemia in cord blood CD34+ cells. · Prieto C, Marschalek R, Kühn A, Bursen A, Bueno C, Menéndez P · Oncotarget 2017
   PMID 29137234DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. A 36-week negative in a xenograft is a bounded observation, not proof of non-leukemogenicity, and cord-blood CD34+ cells are not the fetal-liver cell of origin this project favours for infant t(4;11). Published in Oncotarget.

   ▸Evidence that argues against this (1)

   • Enhanced primary reconstitution without self-renewal on serial transplantation is a negative for transformation, and it is in cord-blood CD34+ cells — neonatal, where this graph records that the fusion consistently fails to transform while an authentic engineered translocation in FETAL liver cells does drive infant-ALL programmes. Oncofusion knowledge graph (internal analysis) 2026
4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

   SIAH1 has DepMap mean gene effect +0.0266 across 1,208 lines with 2.2% of lines depleted and a KMT2A-rearranged minus other-leukemia difference of -0.062 (11 vs 43 lines); SIAH2 has mean +0.054 with 0.17% depleted and a difference of -0.047.

   DepMap CRISPR (Chronos) gene effect via the Open Targets Platform GraphQL API, with KMT2A-rearrangement status from Cellosaurus; per-gene recomputation for this beat · Broad Institute DepMap; Cellosaurus; Open Targets Platform; analysis run by this project · Database and derived analysis 2026
   ▸What this evidence cannot show

   DepMap release version not stated by the API. A knockout screen cannot detect a gene whose function is to degrade an oncoprotein: losing it should if anything help the cell, so this null is the expected result and is uninformative about the biology.

   ▸Evidence that argues against this (1)

   • A mean gene effect of +0.0266 with 2.2% of lines depleted is a null, and a difference of −0.062 across eleven versus 43 lines does not survive being stated as an effect. Oncofusion knowledge graph (internal analysis) 2026
5. E_q91_siah1_loc1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   UniProtKB Q8IUQ4 (SIAH1_HUMAN, primary gene SIAH1, 282 aa, read back live 2026-09-21) lists TWO subcellular location values, 'Cytoplasm' and 'Nucleus', BOTH WITH EMPTY EVIDENCE LISTS, plus a location note - 'Predominantly cytoplasmic. Partially nuclear' - which ALSO has an empty evidence list. No ECO code of any kind appears on any part of this protein's subcellular annotation: no experimental attribution, no curator inference, no rule, no by-similarity source.

   UniProtKB Q8IUQ4 SUBCELLULAR LOCATION comments, evidence codes and feature table, fetched live 2026-09-21 (two location values and a note, all three with empty evidence lists) · UniProt Consortium · UniProtKB 2026
   ▸What this evidence cannot show

   'Established' grades the fact that this is what the record says, not the localisation itself. Tranche 3 found the same empty-evidence form on MAP3K14 and declined to convert it into a NOT APPLICABLE; the same decision is taken here, because SIAH1 is a soluble RING E3 with a real nuclear pool in the primary literature.

   ▸Evidence that argues against this (1)

   • The note is not merely unevidenced, it is contradicted in emphasis by the one direct measurement that exists. 'Predominantly cytoplasmic, partially nuclear' describes a resting distribution; the published experiment describes SIAH-1 REDISTRIBUTING WITHIN THE NUCLEUS TO THE NUCLEAR MATRIX when expressed, together with a mitotic-catastrophe phenotype. A reader taking UniProt's note as the answer would record a predominantly cytoplasmic protein with no subnuclear behaviour, and would be wrong on the only dimension anyone has looked at. Two entries in this tranche have empty-evidence location comments and in both - SMARCB1 is the other - the primary literature is dramatically richer than the database. Bruzzoni-Giovanelli H et al. 1999
6. E_q91_siah1_loc2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   SIAH-1, the human homologue of Drosophila seven in absentia and a p53-p21(Waf-1) inducible gene, blocks growth when stably transfected into the epithelial breast cancer line MCF-7, and the transfectants show a REDISTRIBUTION OF SIAH-1 PROTEIN WITHIN THE NUCLEUS, MORE SPECIFICALLY TO THE NUCLEAR MATRIX, associated with dramatic changes in cell morphology and defective mitosis: multinucleated giant cells with 2-12 nuclei in more than 50% of cells, tubulin spindle disorganisation, defective cytokinesis, and a high frequency of abortive mitotic figures, DNA bridges and persistent intercellular bridges and midbodies.

   SIAH-1 inhibits cell growth by altering the mitotic process. · Bruzzoni-Giovanelli H, Faille A, Linares-Cruz G, Nemani M et al. · Oncogene 1999
   PMID 10597311DOI
   ▸What this evidence cannot show

   1999, one cell line, stable OVER-EXPRESSION, and read from the abstract. Nuclear matrix is an operational fractionation compartment rather than an imaged body. Single source with no retrieved replication in twenty-six years. Not haematopoietic and not fusion-bearing.

   ▸Evidence that argues against this (1)

   • The causal direction is unresolved and the experiment cannot resolve it. SIAH-1 over-expression produces both the nuclear-matrix redistribution AND a mitotic catastrophe, and mitotic catastrophe by itself reorganises the nuclear matrix and the midbody compartment in cells that never touched SIAH-1. So the redistribution may be a readout of the arrest rather than a property of the protein. That matters for how this graph uses the fact: as 'SIAH1 has been seen at the nuclear matrix under forced expression' rather than 'SIAH1 is a nuclear matrix protein'. The midbody detail in the same phenotype is a further reason for caution, because this graph's own WDR5 entry records a centrosome and midbody pool and the two observations could share an over-expression artefact. Lee SB et al. 2012
7. E_q91_siah1_loc3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   A second, independent nuclear pool with its own regulation: B23/nucleophosmin physiologically binds both SIAH1 and GAPDH and DISRUPTS THE SIAH1-GAPDH COMPLEX IN THE NUCLEUS in response to nitrosative stress. S-nitrosylation of B23 at cysteine 275, trans-nitrosylated from GAPDH, dramatically reduces the SIAH1-GAPDH interaction, enhances B23-SIAH1 binding, and mediates neuroprotection by ABROGATING THE E3 LIGASE ACTIVITY OF SIAH1. B23 itself is described as shuttling between nucleolus/nucleoplasm and nucleus/cytoplasm.

   S-nitrosylation of B23/nucleophosmin by GAPDH protects cells from the SIAH1-GAPDH death cascade. · Lee SB, Kim CK, Lee KH, Ahn JY · J Cell Biol 2012
   PMID 23027902DOI
   ▸What this evidence cannot show

   Neuronal and neuroprotective biology - SH-SY5Y cells and mice - with no leukaemia content. It places SIAH1 activity in the nucleus and identifies a nucleolar shuttling partner, but it does not itself assign SIAH1 to the nucleolus and this entry does not either.

   ▸Evidence that argues against this (1)

   • It would be easy to convert 'SIAH1 is regulated by a nucleolar shuttling protein' into a nucleolar assignment for SIAH1, and the bounded query in this entry shows exactly how that error would be manufactured: the only record returned by SIAH1[tiab] AND nucleolar[tiab] is a Parkinson's lncRNA paper where 'nucleolar' is a fragment of the gene name SNHG1, small nucleolar RNA host gene 1. One collision plus one plausible inference is enough to produce a confident wrong compartment, and both were available here. Chen Y et al. 2018
8. E_q91_siah1_loc4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   The SIAH-AFF1 axis has a second publication and it is NOT an independent replication: PMID 21233834 (Leukemia 2011, Pless, Oehm, Knauer, Stauber, Dingermann, Marschalek) reports that the AF4-MLL oncoprotein is activated by Taspase1-mediated hydrolysis into AF4-MLL-N and MLL-C, that these heterodimerise through their FYRN and FYRC domains, and that the resulting high-molecular-weight complex is PROTECTED AGAINST SIAH-mediated degradation. The founding SIAH/AF4 paper this entry already cites, PMID 15221006 (Bursen, Moritz, Gaussmann, Moritz, Dingermann, Marschalek), shares two authors including the senior author.

   The heterodimerization domains of MLL-FYRN and FYRC--are potential target structures in t(4;11) leukemia. · Pless B, Oehm C, Knauer S, Stauber RH et al. · Leukemia 2011
   PMID 21233834DOI
   ▸What this evidence cannot show

   Same laboratory, so this extends rather than replicates the mechanism. Read from the abstract. The system is AF4-MLL, the RECIPROCAL der(4) product, not KMT2A::AFF1.

   ▸Evidence that argues against this (1)

   • This settles one of this entry's own open questions against the entry's hope. The question asked whether any group outside Frankfurt has replicated the SIAH1/SIAH2-AF4 interaction; the only other MLL-scoped SIAH record in PubMed is from the same senior author, so the answer is no, and a mechanism that has stood for twenty-two years on one laboratory's work should be graded accordingly. Two of this graph's arguments about the reciprocal product's stability rest on it. It also adds a structural reason the degradation escape may be indirect - protection arises from high-molecular-weight complex formation after Taspase1 cleavage, i.e. from burial, not necessarily from loss of a degron. Bursen A et al. 2004

• binds →AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing

  binds the AFF1/AF4 N-terminal portion and targets wild-type AFF1 for proteasomal degradation

• related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)

  NOT an interactor of the der(11) product: the AFF1 N-terminus is carried by the reciprocal AFF1::KMT2A, not by KMT2A::AFF1

• related to →SIAH2 — the second AFF1 E3 ligase, reported only alongside SIAH1 and never separated from it

  paralogue pair reported only jointly; no separation-of-function experiment exists

• ← related toRUNX1 — activated by KMT2A::AFF1 and then bound by the reciprocal fusion, in the one published two-oncoprotein circuit in t(4;11)

  both bear on whether the reciprocal der(4) product is functionally important in t(4;11)

• ← depends onWhat 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

  Siah-1a binds the Af4 PXAX-VXP motif at residues 268-290 and degrades it; ~15-fold affinity loss in the robotic mutant

• ← related toThe four AFF4 two-hybrid partners are three rows of one screen, not three experiments - and in that same screen GOLGA2 has 249 partners while AFF4 has five

  the fifth AFF4 two-hybrid partner from the same screen, and the only one already written

• related to →Nuclear matrix and mitotic chromosome scaffold — KMT2A sits with topoisomerase II, in the same place the gene breaks

  SIAH-1 redistributes WITHIN THE NUCLEUS to the NUCLEAR MATRIX on forced expression in MCF-7, with mitotic catastrophe (PMID 10597311) - positive but single-source, over-expression, and causally ambiguous (Q91 2026-09-22)

• related to →The N320-C180 clamp is the only KMT2A module whose disruption has ever been shown to change subnuclear localisation - and the fusion is the disruption

  FYRN/FYRC heterodimerisation of Taspase1-cleaved AF4-MLL forms a high-molecular-weight complex PROTECTED against SIAH-mediated degradation (PMID 21233834) - same laboratory as the founding SIAH/AF4 paper, so not an independent replication (Q91 2026-09-22)