MYH11 — domain architecture and residue coordinates

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.

• The rod interface is generic myosin coiled coil, shared with skeletal and non-muscle myosins; anything that blocks multimerisation through it risks hitting the cytoskeleton broadly, and MYH11 loss-of-function in humans causes familial thoracic aortic aneurysm and dissection.
• Because multimerisation is cooperative and open-ended rather than a fixed dimer, partial inhibition may leave the transforming activity intact, the same objection that applies to the ETV6 SAM polymer.
• This node's central coordinate gap is itself a risk: any figure that draws the fusion junction at a specific MYH11 residue is currently drawing a number nobody in this pass could source.

1. What is the first retained MYH11 residue in the type A CBFB::MYH11 fusion, in canonical 1,972 aa numbering? This is a tractable question that would be answered by mapping the published type A cDNA junction onto the canonical coding sequence, and it is the main gap in this node.
2. Where exactly is the assembly competence domain in MYH11 numbering? Its boundaries were described as uncertain even by the group that mapped it functionally, and it is the one part of the rod that matters.
3. The four transcript types differ in how much rod they include. Does rod length correlate with multimer size, with degree of core-binding-factor inhibition, or with anything clinical? That is a natural dose-response experiment that the patient population has already run.
4. Is there any structural information on the MYH11 rod at all? This node assumed there was and there is not. A cryo-EM structure of even a rod fragment containing the assembly competence domain would convert the central mechanistic claim of CBFB::MYH11 from deletion mapping into coordinates, and would finally give the ACD a residue range.

1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   MYH11 canonical is 1,972 residues with a myosin N-terminal SH3-like domain at 31-81, a myosin motor domain at 85-783, an IQ motif at 786-815, a coiled coil at 844-1934 and a non-helical C-terminal tailpiece at 1935-1972.

   UniProtKB entry P35749 (MYH11_HUMAN): reviewed sequence, length and positional feature annotations · UniProt Consortium · UniProtKB/Swiss-Prot 2026
   ▸What this evidence cannot show

   The coiled-coil annotation is a single 1,091-residue feature; real myosin rods contain skip residues and local discontinuities that a single interval hides. Isoform lengths were retrieved directly and are 1972, 1979, 1945 and 1938 for P35749-1 through -4.

   ▸Evidence that argues against this (2)

   • UniProt's single 844-1934 coiled-coil feature is contradicted in form, not just in boundary, by the structural classifications: CATH-Gene3D and SUPERFAMILY tile the same region as about seven separate myosin-rod fragments, and Pfam PF01576 gives 848-1928. UniProt's own tailpiece cut at 1934/1935 falls inside a CATH-assigned unit spanning 1901-1944, so the rod/tailpiece boundary this node draws is not supported by the structural classification it is drawn against. InterPro Consortium; EMBL-EBI 2026
   • The IQ motif is the fate-uncertain entry in this node and its C-terminal boundary is itself disputed: InterPro IPR000048 and SMART SM00015 both end it at 807, eight residues before UniProt's 815. Since the node already flags the IQ motif as the domain that could fall on either side of some CBFB::MYH11 breakpoints, an eight-residue disagreement on its own end is not negligible. InterPro Consortium; EMBL-EBI 2026
2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   A structure of smooth-muscle myosin II (PDB 9FU2) maps to MYH11 residues 1-1972, i.e. the full-length chain.

   PDB 9FU2 'Smooth Muscle Myosin II in complex with MT-228' (title retrieved and verified at RCSB; released 2025) · RCSB Protein Data Bank (entry metadata retrieved and title verified) · RCSB PDB 2026
   ▸What this evidence cannot show

   Full-chain UniProt mapping does not mean every residue is resolved; long coiled-coil regions in such structures are frequently modelled at low resolution or as poly-alanine.

   ▸Evidence that argues against this (1)

   • CORRECTION, and the reason the body of this node was rewritten. 9FU2's SIFTS mapping to P35749 1-1972 is the DEPOSITED SEQUENCE, not the observed model. Residues 787-1972 carry no coordinates at all. The rod was not visualised; the ordered region stops within a residue or two of the IQ motif. The claim that 'the motor and at least part of the rod have been visualised in one assembly' was wrong and has been withdrawn. RCSB Protein Data Bank (entry and instance metadata retrieved and read back) 2026
3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

   The inv(16) fusion connects the first 165 amino acids of CBFB to the tail region of smooth-muscle myosin heavy chain, and the repeated coiled coil of SMMHC was proposed as the source of fusion-protein dimerisation.

   Fusion between transcription factor CBF beta/PEBP2 beta and a myosin heavy chain in acute myeloid leukemia · Liu P, Tarle SA, Hajra A, et al. · Science 1993
   PMID 8351518DOI
   ▸What this evidence cannot show

   The MYH11 contribution is described as 'the tail region' without a residue number; this is the origin of the residue-level gap flagged in this node.

   ▸Evidence that argues against this (1)

   • The dominant-negative phenotype has an alternative, measured explanation that needs no dimerisation. Isothermal titration calorimetry shows CBFbeta-SMMHC binds the RUNX1 Runt domain with HIGHER affinity than native CBFbeta does, with NMR mapping interactions in both the CBFbeta portion and the SMMHC coiled coil. Higher-affinity sequestration of RUNX1 by itself accounts for the dominant-negative phenotype of the CBFB-MYH11 knock-in, so the 1993 dimerisation proposal is one hypothesis among at least two - and that paper reports no oligomerisation measurement of its own. Lukasik SM et al. 2002
4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   A CBFbeta-SMMHC variant with the assembly competence domain deleted does not multimerise, does not efficiently bind co-repressors, and fails to slow myeloid progenitor proliferation, unlike the intact fusion.

   Identification of a region on the outer surface of the CBFbeta-SMMHC myeloid oncoprotein assembly competence domain critical for multimerization · Zhang L, D'Costa J, Kummalue T, Civin CI, Friedman AD · Oncogene 2006
   PMID 16767164DOI
   ▸What this evidence cannot show

   The 2006 paper defines the ACD functionally by deletion and helix-surface mutagenesis rather than giving a crisp residue range in MYH11 canonical numbering; it also notes that 'divergence at the boundaries of the ACD may reflect quantitative differences between in vitro and in vivo filament assembly', i.e. the ACD boundary is itself unsettled. DOWNGRADED strong -> moderate on 2026-09-19: the multimerisation readouts are confounded with nuclear localisation (PMID 12417730), so the causal step cannot be assigned from these experiments.

   ▸Evidence that argues against this (1)

   • The ACD deletion also moves the protein out of the nucleus, which confounds every downstream readout. The same laboratory reported four years earlier that CBFbeta-SMMHC(deltaACD) and the short CBFbeta-SMMHC(83C) variant are PREDOMINANTLY CYTOPLASMIC while wild-type and CBFbeta-SMMHC(149C) are mostly nuclear, and that forcing deltaACD into the nucleus with an SV40 NLS let a significant fraction enter but still only mildly inhibited CBF activities. They also showed that directing the ACD to CBF target genes by fusing it to the AML1 DNA-binding domain, or to full-length AML1, produced no effect on Ba/F3 proliferation. Loss of function on ACD deletion is therefore at least partly loss of nuclear retention, and the ACD is not sufficient on its own. Kummalue T et al. 2002
5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   Mutating the outer surface of individual ACD helices reduces multimerisation, with helices D and E having the largest effect, and each variant's ability to slow myeloid progenitor proliferation largely parallels its multimerisation.

   Identification of a region on the outer surface of the CBFbeta-SMMHC myeloid oncoprotein assembly competence domain critical for multimerization · Zhang L, D'Costa J, Kummalue T, Civin CI, Friedman AD · Oncogene 2006
   PMID 16767164DOI
   ▸What this evidence cannot show

   Multimerisation was assayed in low ionic strength in vitro and inferred in vivo from nuclear localisation, both indirect readouts. Cited alongside the companion functional study (PMID 15815715). DOWNGRADED strong -> moderate on 2026-09-19: the multimerisation readouts are confounded with nuclear localisation (PMID 12417730), so the causal step cannot be assigned from these experiments.

   ▸Evidence that argues against this (1)

   • Parallels is compatible with a shared upstream cause rather than causation, and the same laboratory supplied that cause. The ACD's role was shown to be mediating multimerisation IN THE NUCLEUS, with ACD-deficient variants predominantly cytoplasmic; the later paper itself inferred in-vivo multimerisation from nuclear localisation. A helix mutation that weakens multimerisation will also weaken nuclear retention, so a correlation between multimerisation score and proliferation slowing is expected under either model, and a low-ionic-strength in-vitro multimerisation assay does not break the tie. Kummalue T et al. 2002
6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

   Four distinct CBFB-MYH11 chimeric transcripts exist, differing in the MYH11 fusion point while all splicing to CBFB nucleotide 495.

   Detection of fusion transcripts generated by the inversion 16 chromosome in acute myelogenous leukemia · Claxton DF, Liu P, Hsu HB, et al. · Blood 1994
   PMID 8142642
   ▸What this evidence cannot show

   RT-PCR on 37 cases; MYH11 fusion points are reported at the cDNA level, not as protein residues. No DOI indexed; cited by PMID.

   ▸Evidence that argues against this (1)

   • The count was low by a factor of more than two, and the missing variants are diagnostically invisible. A later report states that ten different in-frame CBFB-MYH11 fusion transcripts had by then been identified, and describes an eleventh whose MYH11 junction sits at cDNA position 2134 - OUTSIDE the region amplified by the commonly used primers - so all previously established RT-PCR protocols may miss it, and the authors propose a two-step assay to catch all of them. Any residue-level MYH11 breakpoint map built on the 1994 four-transcript set is truncated at both the low- and high-frequency ends. Van der Reijden BA et al. 2001
7. E7established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   An AlphaFold model AF-P35749-F1 covers all 1,972 residues.

   AlphaFold Protein Structure Database entry for P35749 (model AF-P35749-F1) · Jumper J, Varadi M, et al.; EMBL-EBI/DeepMind · AlphaFold DB 2026
   ▸What this evidence cannot show

   AlphaFold models of very long coiled coils are unreliable in their global geometry even when locally confident; the predicted rod should not be used to measure anything.

   ▸Evidence that argues against this (1)

   • Coverage is complete but confidence is mostly mediocre, and the mediocre part is exactly the fusion-relevant part. Per-residue pLDDT recomputed from the v6 model gives only 383 of 1,972 residues (19%) at or above 90, 939 (48%) in the 70-90 band, 572 (29%) in the 50-70 band and a mean of 76.3 - so roughly a third of the chain sits at or below the 70 threshold, and the long rod carrying the assembly competence domain is dominated by the 50-70 band, where AlphaFold's own guidance says the prediction should be treated with caution. There is exactly one experimental structure for P35749 in the whole PDB (9FU2) and it observes nothing past residue 786. AlphaFold DB / EMBL-EBI; PDBe SIFTS 2026
8. E8established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   In PDB 9FU2 the deposited MYH11 chain is full length (1,972 residues in SEQRES) but residues 787-1972 are UNOBSERVED, together with 1-31, 50-51, 63-70, 202-211 and 629-650; the coordinates therefore cover only the SH3-like domain, the motor domain and the first residue or two of the IQ motif, and contain no part of the coiled-coil rod.

   PDB 9FU2 'Smooth Muscle Myosin II in complex with MT-228': observed-residue ranges from the RCSB polymer entity instance record for chain A · RCSB Protein Data Bank (entry and instance metadata retrieved and read back) · RCSB PDB 2026
   ▸What this evidence cannot show

   Read from the RCSB instance record's UNOBSERVED_RESIDUE_XYZ feature positions on 2026-09-19, i.e. the deposited model, not the paper (database record; no full text read). SEQRES-length SIFTS mappings are routinely mistaken for observed coverage and that is exactly the error this item corrects.

   ▸Evidence that argues against this (1)

   • This item is right, and the counter-evidence is that a second EBI resource publishes the opposite impression for the same entry. The PDBe SIFTS best-structures endpoint for P35749 returns exactly one structure, 9FU2, with unp_start 1 and unp_end 1972 at 2.581 A - advertising the entry as covering the complete 1,972-residue protein, with no indication that 1,186 of those residues (60%) have no coordinates. Anyone importing MYH11 structural coverage from SIFTS rather than from the RCSB instance record will read full-length coverage, and that disagreement between two records of the same deposition is the hazard. PDBe / SIFTS et al. 2026
9. E9strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

   Member databases disagree with UniProt on every MYH11 boundary this node stores: the myosin N-terminal SH3-like domain is 24-81 in Pfam PF02736/InterPro IPR004009 and only 31-64 in SUPERFAMILY SSF50084 against UniProt's 31-81; the motor domain is 79-784 in SMART/InterPro and 99-771 in CDD cd14921 against UniProt's 85-783; the IQ motif ends at 807 in InterPro IPR000048 and SMART SM00015 against UniProt's 815; and the rod, which UniProt draws as one 844-1934 feature, is modelled by Pfam PF01576 as 848-1928 and by CATH-Gene3D and SUPERFAMILY as roughly seven separate segments (843-964, 963-1078, 1071-1157, 1243-1355, 1357-1457, 1521-1634 and 1901-1944).

   InterPro member-database domain matches for UniProt P35749 (Pfam, PROSITE, SMART, CDD, CATH-Gene3D, SUPERFAMILY) · InterPro Consortium; EMBL-EBI · InterPro 2026
   ▸What this evidence cannot show

   Profile-model match extents, not measurements. Retrieved 2026-09-19 (database record). The fragmented CATH/SUPERFAMILY rod assignments are a consequence of how those methods tile long coiled coils and should not be read as evidence of seven discrete folds; they are evidence that the single 844-1934 bar is a convenience.

   ▸Evidence that argues against this (1)

   • The disagreement is real but unarbitrable, and that bounds what the item can be used for. P35749 has exactly ONE entry in the entire PDB - 9FU2, X-ray, 2.581 A - and converting its UNOBSERVED_RESIDUE_XYZ features to author numbering shows the deposited coordinates stop at residue 786 (unobserved: 1-31, 50-51, 63-70, 202-211, 629-650 and 787-1972). No experimental structure exists anywhere for residues 787-1972, so not one of the competing rod boundaries (844, 848, 843, 1928, 1934, 1944, or the seven-segment CATH/SUPERFAMILY tiling) has ever been checked against coordinates. This should be read as 'no source knows', not as 'sources disagree and one is right'. PDBe / SIFTS et al. 2026

• related to →CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)

  Counter-evidence added cycle 41: the multimerisation-causes-inhibition chain is confounded with nuclear localisation. ACD-deficient CBFbeta-SMMHC variants are predominantly CYTOPLASMIC, and forcing deltaACD into the nucleus with an SV40 NLS still only mildly inhibited CBF activities (Kummalue 2002, PMID 12417730). Items E4 and E5 downgraded strong -> moderate.

• refutes →Hypothesis — forced-oligomerisation interfaces are the mutation-resistant drug targets the fusion field has been ignoring

  Counter-evidence for an oligomerisation-interface target premise: CBFbeta-SMMHC's dominant-negative phenotype has a measured alternative explanation that needs no oligomer - it binds the RUNX1 Runt domain with HIGHER affinity than native CBFbeta by isothermal titration calorimetry (Lukasik 2002, PMID 12172539). And no experimental structure exists anywhere for MYH11 residues 787-1972, so no rod boundary has ever been checked against coordinates.

• related to →CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)

  Correction (RCSB observed-residue record for 9FU2): the rod that makes this fusion oligomerise has never been structurally resolved. 9FU2 is full length in SEQRES but carries no coordinates past residue 786, so everything this graph says about the MYH11 coiled coil and the assembly competence domain inside it rests on prediction and deletion mapping, not on atoms.

• related to →CBFB — domain architecture and residue coordinates

  Counter-evidence on both sides of the same junction: the MYH11 IQ motif's C-terminal boundary is 807 (InterPro/SMART) or 815 (UniProt), and the CBFB fold is ordered to residue ~157 in deposited structures rather than ending at the stored 141. The junction at CBFB 165 is therefore closer to ordered structure on the CBFB side, and less certainly placed on the MYH11 side, than either entry drew.

• ← fuses withCBFB — domain architecture and residue coordinates

  CBFB::MYH11, inv(16)

• part of →CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)

  3-prime partner domain architecture; the coiled-coil rod is the retained multimerisation element

• supports →Hypothesis — forced-oligomerisation interfaces are the mutation-resistant drug targets the fusion field has been ignoring

  the MYH11 rod multimerises the fusion and sequesters RUNX1

• fuses with →CBFB — domain architecture and residue coordinates

  inv(16)(p13q22) / t(16;16)(p13;q22)

• supports →Mechanism class 2 — transcription-factor corruption (RUNX1::RUNX1T1, PML::RARA, ETV6::RUNX1, CBFB::MYH11) — a balance, not a pure dominant negative

  multimerisation through the rod's assembly competence domain is required for the fusion to inhibit core-binding factor

• tests →Hypothesis — forced-oligomerisation interfaces are the mutation-resistant drug targets the fusion field has been ignoring

  ACD helix-surface mutants separate multimerisation from corepressor binding, a direct test of whether the interface is the right target

• supports →Do different breaks give different biology? A verdict on partner identity vs breakpoint position

  four CBFB::MYH11 transcript types include different lengths of rod, a natural dose-response in multimerisation capacity

• part of →CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)

  3' partner domain coordinates