Oncofusion

start / graph / gene-ss18

Proteinmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q47-swi-snf

SS18 - a SWI/SNF subunit that is itself a fusion partner in another cancer, and a lineage-independent dependency in KMT2A-rearranged lines

also called SSXT · SYT · SS18L1 paralogue · SSXT_HUMAN

SS18 is a small subunit shared by two flavours of the SWI/SNF machine. It is also the gene that fuses to SSX in synovial sarcoma, producing a fusion protein that lives inside the machine itself - which is the cleanest example anywhere of a fusion oncoprotein hijacking chromatin remodelling, and a useful contrast with how KMT2A fusions work.

The same thing, for a scientist

SS18 (Q15532, 418 aa) is a shared subunit of cBAF and ncBAF [E1][E2]. It matters here for two independent reasons. **First, dependency**: rank 224/17,728 in the KMT2A-rearranged selectivity scan (difference -0.2124, pan-essentiality +0.0395) and depleted in both lineage halves (myeloid z -0.92, B-lineage -1.27) - one of four SWI/SNF genes to survive the split [E3][E4]. **Second, as the comparator case**: in synovial sarcoma, SS18-SSX - just 78 amino acids of SSX fused to SS18 - competes for assembly with wild-type SS18, produces an altered BAF complex lacking BAF47, binds the SOX2 locus, reverses polycomb repression and activates SOX2; restoring wild-type SS18 reassembles normal complexes, retargets them away from SOX2 and stops proliferation [E5]. **No KMT2A fusion partner is a SWI/SNF subunit**, so the KMT2A mechanism cannot work this way, and SS18's role in this beat is as the boundary of the analogy rather than as a model for it [E6]. `SS18 AND (MLL OR KMT2A)` returns 3 records [E7].

What it looks like
loading AF-Q15532-F1…
AlphaFold AF-Q15532-F1 ↗
colourshape

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.

very high (90+)confident (70–90)low (50–70)very low (<50)
I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper
Two ways a fusion can meet BAF, and KMT2A does neither

The fusion-oncoprotein literature contains exactly two demonstrated mechanisms by which a fusion engages mSWI/SNF:

  1. Be a subunit. SS18-SSX is incorporated into BAF in place of wild-type SS18 and retargets the whole complex 5.
  2. Recruit it through a prion-like domain. EWS-FLI1 recruits BAF to tumour-specific GGAA-microsatellite enhancers; this is a neomorphic property of the fusion compared with wild-type FLI1 and depends on tyrosines required for EWSR1 prion-like-domain phase behaviour; fusing short EWSR1 fragments to FLI1 is sufficient to recapitulate BAF retargeting 6.

KMT2A fusions do neither on any published evidence. No KMT2A partner is a BAF subunit, and no KMT2A fusion has been shown to carry BAF to chromatin: BAF AND (MLL-AF4 OR MLL-AF9 OR KMT2A-AFF1) AND immunoprecipitation returns 0 with operands checked 8. That is either a genuine mechanistic difference or an unlooked-for possibility, and this beat cannot distinguish them - but it can say which.

Condensates in two diseases, neither of them this one (Q87, 2026-09-21)

Subcellular. Nucleus - but UniProt Q15532's only location term carries ECO:0000305, a curator inference, with no experimental citation. Same flag as SMARCD2 in rule 91, and unusual for a BAF subunit. HPA v25.1 IF, reliability Approved: nucleoplasm main, cytosol additional.

Subnuclear: tyrosine-IDR-driven nuclear condensates - and say which disease. Mouse ESCs, pluripotent-to-somatic transition (PMID 34215745). Synovial sarcoma, SS18 and SS18-SSX, QPGY-driven LLPS recruiting BRG1 (PMID 35585082). Neither uses calibrated half-bleaching, so 'phase separation' is the model, not the measurement.

The internal counter-argument, from the ESC paper itself. The IDR alone forms condensates indistinguishable from wild type and does not rescue function. Condensate formation is not sufficient.

Cell type and state. B-lymphoid progenitors, via MEF2D-SS18 in B-ALL: nuclear by IF, and causing B-cell developmental arrest at the CLP-to-CD19-positive step (PMID 29408457). That nuclear signal is at least as attributable to the MEF2D half as to SS18.

Shown for fusion: SS18-SSX (synovial sarcoma) and MEF2D-SS18 (B-ALL). NOT a KMT2A fusion - the KMT2A-scoped localisation query returns 0.

Facts and figures
uniprotQ15532
gene symbolSS18
protein nameProtein SSXT
length AA418
alphafold IDAF-Q15532-F1
accession resolutionUniProt REST gene_exact:SS18 AND organism_id:9606 AND reviewed:true, run 2026-09-20; one reviewed row; gene_primary and length read back from the same record.
complexshared between cBAF and ncBAF
classificationlineage-independent KMT2A-rearranged dependency; separately, the SWI/SNF subunit that is itself a fusion partner in synovial sarcoma
subnuclear localizationMICROSCOPIC NUCLEAR CONDENSATES driven by a C-terminal tyrosine-rich IDR (the QPGY domain). Shown in MOUSE EMBRYONIC STEM CELLS for wild-type SS18 during the pluripotent-to-somatic transition, where tyrosine mutation abolishes condensates and rescue but the IDR alone forms condensates WITHOUT rescuing (PMID 34215745), and in SYNOVIAL SARCOMA for SS18 and SS18-SSX, where QPGY self-association drives LLPS and recruits BRG1 (PMID 35585082). Neither study uses calibrated half-bleaching, so 'condensate' is what was imaged and 'phase separation' is the model. The one B-lineage-leukaemia result is a plain nuclear IF on the MEF2D-SS18 fusion (PMID 29408457), not a condensate measurement. No KMT2A fusion anywhere.
cell stateTHE PLURIPOTENT-TO-SOMATIC TRANSITION is the state SS18 was identified as regulating, by CRISPR-based whole-genome screening, and SS18 mediates BAF assembly through phase separation to control it (PMID 34215745). The leukaemia-relevant state statement is different and comes from the fusion: MEF2D-SS18 causes B-CELL DEVELOPMENTAL ARREST at the common lymphoid progenitor to CD19-positive B-cell step (PMID 29408457) - a differentiation block, which is the state this graph cares about most, reached through a non-KMT2A fusion.
shown for fusionNOT A KMT2A FUSION - and saying which fusion is the whole point here. Measured 2026-09-21: (SS18[tiab] OR SYT[tiab] OR SSXT[tiab]) AND the eight-term KMT2A/MLL fusion clause AND the ten-term localisation clause returns 0 records. What DOES exist is localisation in two OTHER fusions: SS18-SSX in SYNOVIAL SARCOMA (LLPS and BRG1 recruitment, PMID 35585082) and MEF2D-SS18 in B-ALL (nuclear by IF, PMID 29408457). Recording a bare NO would collapse 'nobody has looked in any fusion' with 'it was shown in two fusions, neither of them this one', and those are different states of knowledge.
paralogue
{
  "gene": "SS18L1",
  "note": "neural BAF (CREST); classified a myeloid-identity artefact in this beat's lineage split"
}
depmap
{
  "rank": 224,
  "of": 17728,
  "percentile": 1.264,
  "difference": -0.2124,
  "mean_kmt2a_r": -0.2282,
  "mean_other_leukaemia": -0.0158,
  "pan_essentiality_all_lines": 0.0395,
  "lineage_split_z_myeloid": -0.92,
  "lineage_split_z_b_lineage": -1.27,
  "verdict": "lineage-independent"
}
fusion elsewhere
{
  "fusion": "SS18::SSX1/SSX2",
  "disease": "synovial sarcoma",
  "mechanism": "fusion is incorporated into BAF in place of wild-type SS18, evicts BAF47, activates SOX2"
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.2282,
  "chronos_mean_other_leukaemia": -0.0158,
  "difference": -0.2124,
  "pan_essentiality_all_lines": 0.0395,
  "selectivity_rank": 224,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
localization
{
  "subcellular": "NUCLEUS - and UniProt's evidence for it is only ECO:0000305, a CURATOR INFERENCE, not an experimental code. Q15532 carries one location term, 'Nucleus', with no experimental citation behind it. This is the same flag raised in rule 91 item 4 for SMARCD2, and it is unusual for a BAF subunit. HPA v25.1 IF, reliability Approved, gives Nucleoplasm main with Cytosol additional, and is the better-founded of the two. The most direct experimental nuclear call in the literature is on a FUSION rather than on wild-type SS18: immunofluorescent staining shows nuclear localisation of MEF2D-SS18 in B-ALL models (PMID 29408457).",
  "subnuclear": "MICROSCOPIC NUCLEAR CONDENSATES formed through a C-terminal tyrosine-rich intrinsically disordered region, reported twice in two different diseases, and the assays used matter. (1) In MOUSE EMBRYONIC STEM CELLS during the pluripotent-to-somatic transition, SS18 forms microscopic nuclear condensates through a C-terminal tyrosine-rich IDR; mutating the tyrosines abolishes both condensate formation and rescue of the SS18-null defect, and SS18 is reported to mediate BAF assembly through phase separation. Notably the IDR ALONE forms condensates indistinguishable from wild type and does NOT rescue - so condensate formation is demonstrably not sufficient for function (PMID 34215745). (2) In SYNOVIAL SARCOMA, self-association of the QPGY domain drives LLPS of SS18 and of SS18-SSX and recruits BRG1 into the condensates; tyrosines in the QPGY domain are decisive, and perturbing either LLPS or BRG1 binding impairs NIH3T3 transformation by SS18-SSX (PMID 35585082).",
  "which_disease_each_result_came_from": "This matters because SS18's literature is dominated by synovial sarcoma. The pluripotency condensate result is mouse ESC developmental biology (PMID 34215745). The LLPS and BRG1-recruitment result is synovial sarcoma biochemistry with an NIH3T3 fibroblast transformation readout (PMID 35585082). The one B-LINEAGE LEUKAEMIA localisation result is on MEF2D-SS18, an SS18 fusion found in B-ALL, and it is a plain nuclear-localisation IF, not a condensate measurement (PMID 29408457). NONE of the three involves a KMT2A fusion.",
  "condensate_evidence_standard": "Per this graph's standing condensate caution: neither paper uses calibrated half-bleaching to measure internal mixing, which is the only model-free assay that separates liquid-liquid phase separation from binding to spatially clustered sites. Both work with tagged, overexpressed or in-vitro-reconstituted protein. 'Condensate' is therefore the correct descriptive word for what was imaged, and 'phase separation' is a model the data are consistent with rather than a demonstrated mechanism - and the ESC paper's own result that the IDR forms condensates without rescuing function is the strongest internal argument for keeping those two claims apart.",
  "how_shown": "CRISPR whole-genome screening plus fluorescence imaging of tagged SS18 with tyrosine-to-other IDR mutants and domain-swap rescues in mouse ESCs (PMID 34215745); cryo-EM/crystallographic structures of SS18/BRG1 and SNF11/SNF2 subcomplexes plus in-vitro and cellular LLPS assays and NIH3T3 transformation (PMID 35585082); immunofluorescent staining of MEF2D-SS18 in B-ALL models (PMID 29408457)",
  "node": "localization-condensate-evidence-standards"
}
cell type
{
  "haematopoietic_answer": "B-LYMPHOID PROGENITORS, and only through a different fusion. MEF2D-SS18 is one of the MEF2D-family fusions defining a B-ALL subtype with relatively worse survival; it localises to the nucleus by IF, inhibits differentiation of mouse common lymphoid progenitors into CD19-positive B cells, and produces B-cell developmental arrest in a mouse model (PMID 29408457). That is a genuinely haematopoietic, genuinely B-lineage answer for SS18 sequence - but it is SS18 as a fusion PARTNER, not wild-type SS18, and the disease is MEF2D-rearranged B-ALL, not KMT2A-rearranged leukaemia.",
  "the_other_systems": "mouse embryonic stem cells (condensates); synovial sarcoma and NIH3T3 fibroblasts (LLPS, BRG1).",
  "expression_atlas": "HPA v25.1: RNA tissue specificity Low; IF reliability Approved."
}
How this could bite us back
  • The synovial sarcoma precedent shows a BAF-directed intervention can work by RESTORING normal complex composition rather than by inhibiting anything - a mode no agent in the leukaemia pipeline pursues, and a reason inhibition may be the wrong verb.
  • No SS18-directed agent exists; the dependency is currently only actionable through BRD9 or the shared ATPase, both of which hit more than SS18.
  • If the KMT2A-rearranged SS18 requirement runs through cBAF, then the ncBAF interpretation of this beat's dependency data is wrong and a BRD9 degrader would not capture it.
Still open
  1. Has anyone tested whether a KMT2A fusion's low-complexity partner region recruits BAF the way EWSR1's prion-like domain does? The experiment exists (Boulay's EWSR1-fragment fusions) and has never been run with an AFF1 or MLLT3 fragment.
  2. Is SS18's KMT2A-rearranged dependency the cBAF pool, the ncBAF pool, or both? SS18 is the only subunit in this beat that could be used to separate them by complex-selective immunoprecipitation.
  3. Why does SS18 score while its paralogue SS18L1 reads as a lineage artefact?
Every source behind this entry (10)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    SS18 is UniProt Q15532, 418 aa, primary gene SS18, synonyms SSXT SYT; entry name SSXT_HUMAN.

    UniProtKB Q15532 (SSXT_HUMAN) · UniProt Consortium · UniProtKB 2026
    What this evidence cannot show

    gene_exact:SS18 AND organism_id:9606 AND reviewed:true, run 2026-09-20; one reviewed row; gene_primary and length read back from the fetched record. The synonym SYT collides with the synaptotagmin gene family in literature searches.

    Evidence that argues against this (1)
    • SS18L1 (CREST) is a paralogue occupying the same slot in neural BAF; a dependency attributed to SS18 could belong to the slot. SS18L1 is a myeloid-identity artefact in this beat's lineage split, so they behave differently here. Oncofusion knowledge graph et al. 2026
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    SS18 is a component of both canonical BAF and the GBAF/ncBAF subcomplex.

    PMID 29374058DOI
    What this evidence cannot show

    Composition from density gradient and IP in prostate cancer lines.

    Evidence that argues against this (1)
    • Because SS18 is shared between cBAF and ncBAF, its dependency cannot be attributed to ncBAF on composition alone - it is consistent with either. Michel BC et al. 2018
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    SS18 ranks 224 of 17,728 for KMT2A-rearranged selectivity with difference -0.2124 (KMT2A-r -0.2282, other leukaemia -0.0158, pan-essentiality +0.0395).

    Genome-wide KMT2A-rearranged selectivity scan, 17,728 genes (DepMap Chronos via Open Targets) · Oncofusion knowledge graph, analyses/depmap-kmt2a-selective-scan-full/run.py · DepMap Chronos gene effect via Open Targets GraphQL 2026
    What this evidence cannot show

    11 index against 43 comparator lines; pan-essentiality above zero means the average line does not require SS18, which both makes the window attractive and inflates the z-derived rank relative to the raw effect.

    Evidence that argues against this (1)
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    SS18 is depleted in both lineage halves of the KMT2A-rearranged arm (myeloid z -0.92, B-lineage -1.27) and is one of four SWI/SNF panel genes to survive the split.

    SWI/SNF panel lineage-split DepMap dependency scan, 28 genes · Oncofusion knowledge graph, analyses/swi-snf-lineage-split/run.py · DepMap Chronos gene effect via Open Targets GraphQL 2026
    What this evidence cannot show

    Eight myeloid and five B-lineage index lines, none infant; both reference myeloid lines also show moderate depletion (OCI-AML3 -0.87, U-937 -0.81), so the rearrangement association is weaker than for BICRA.

    Evidence that argues against this (1)
    • SS18's paralogue SS18L1 is classified a myeloid-identity artefact by the same test, so the slot as a whole behaves inconsistently and the SS18 result should not be generalised to 'the SS18 slot'. Oncofusion knowledge graph et al. 2026
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In synovial sarcoma, SS18-SSX - 78 amino acids of SSX fused to SS18 - competes for assembly with wild-type SS18, forming an altered BAF complex lacking BAF47 that binds the SOX2 locus and reverses polycomb repression; raising wild-type SS18 reassembles normal complexes, retargets them away from SOX2 and stops proliferation, and the mechanism depends on only two amino acids of SSX.

    PMID 23540691DOI
    What this evidence cannot show

    Synovial sarcoma cell systems; the wild-type SS18 rescue is an over-expression experiment, and 'depends on only two amino acids' is a mapping result in that construct.

    Evidence that argues against this (1)
    • This mechanism requires the fusion to be a complex subunit. It cannot transfer to KMT2A fusions, whose partners are elongation and chromatin-reader proteins, so the analogy is structural rather than mechanistic. Boulay G et al. 2017
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    EWS-FLI1 recruits the BAF complex to tumour-specific enhancers as a neomorphic property compared with wild-type FLI1, dependent on tyrosines required for EWSR1 prion-like-domain phase transitions, and fusing short EWSR1 fragments to FLI1 is sufficient to recapitulate BAF retargeting.

    Cancer-Specific Retargeting of BAF Complexes by a Prion-like Domain. · Boulay G, Sandoval GJ, Riggi N, Iyer S, et al. · Cell 2017
    PMID 28844694DOI
    What this evidence cannot show

    Ewing sarcoma models; the phase-transition attribution rests on tyrosine mutants, and this project's standing caution applies - the assays that support phase separation generally cannot distinguish it from binding to spatially clustered sites.

    Evidence that argues against this (1)
    • A prion-like-domain mechanism predicts that any fusion with such a domain could retarget BAF; KMT2A fusion partners AFF1 and MLLT3 are low-complexity and intrinsically disordered, so the absence of a reported BAF interaction in KMT2A may reflect absence of looking rather than absence of the phenomenon. Shi J et al. 2013
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    `SS18 AND (MLL OR KMT2A)` returns 3 PubMed records and `SS18 AND leukemia` returns 13.

    What this evidence cannot show

    Bounded absence. Bounded to PubMed on 2026-09-20 over SS18 only; the aliases SSXT and SYT were deliberately not searched - SYT collides with synaptotagmin and would inflate rather than recall. The three records were not fetched, so this bounds the count and not their content.

    Evidence that argues against this (1)
    • SS18 appears in the KMT2A-rearranged dependency data without appearing in the KMT2A literature, which is the same pattern as BICRA and is the reason this beat exists. Oncofusion knowledge graph et al. 2026
  8. E8moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    No published experiment places a BAF subunit physically on a KMT2A fusion protein: `BAF AND (MLL-AF4 OR MLL-AF9 OR KMT2A-AFF1) AND immunoprecipitation` returns 0 with operands and a positive control checked.

    What this evidence cannot show

    Bounded absence. Bounded to PubMed on 2026-09-20 over that term set. It cannot see a co-IP reported only in a figure of a paper whose abstract frames the work at complex level (rule 64), nor an unpublished negative. `SMARCA4 AND immunoprecipitation AND leukemia` returns 10 records which were not individually title-checked.

    Evidence that argues against this (1)
    • The graph's own interactome gap analysis shows IntAct carries no KMT2A-SMARCA4 rows at all, which is consistent; but IntAct curation of a 2013-2015 literature is itself incomplete. Mo X et al. 2022
  9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Wild-type SS18 forms microscopic condensates in nuclei through a C-terminal intrinsically disordered region rich in tyrosine, in MOUSE EMBRYONIC STEM CELLS, where SS18 was identified by CRISPR whole-genome screening as a regulator of the pluripotent-to-somatic transition. Mutating the tyrosines abolishes both condensate formation and rescue of the SS18-null defect; the N-terminal 70 amino acids are separately required, through interaction with the BAF complex, and remain functional when swapped onto unrelated IDRs or onto an artificial 24-tyrosine polypeptide.

    SS18 regulates pluripotent-somatic transition through phase separation. · Kuang J, Zhai Z, Li P, Shi R, Guo W, Yao Y, et al. · Nat Commun 2021
    PMID 34215745DOI
    What this evidence cannot show

    MOUSE EMBRYONIC STEM CELLS - not a leukaemia, not a sarcoma and not a haematopoietic cell - with tagged protein. No calibrated half-bleaching was performed, so 'phase separation' is the authors' model rather than a measurement that excludes binding to spatially clustered sites; this graph's standing condensate caution applies in full.

    Evidence that argues against this (1)
    • This paper contains its own best objection to the condensate story and it should be quoted whenever the result is cited: the IDR ALONE forms condensates INDISTINGUISHABLE from wild-type SS18 and does NOT rescue the null defect. So condensate formation is demonstrably not sufficient for function, and the functional determinant is the N-terminal 70 amino acids that bind BAF - which remain functional grafted onto an artificial 24-tyrosine polypeptide. The synovial sarcoma work reaches the same shape from the other side: perturbing EITHER LLPS OR BRG1 binding impairs transformation, i.e. the condensate and the complex are both required and neither is the mechanism on its own. Any proposal built on 'SS18 works by phase separation' is asserting more than either paper shows. Cheng Y et al. 2022
  10. E10moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The one B-lineage leukaemia localisation result involving SS18 sequence is on a fusion, and it is not a KMT2A fusion: immunofluorescent staining shows nuclear localisation of MEF2D-SS18, one of the MEF2D-family fusions defining a B-ALL subtype with relatively worse survival. MEF2D-SS18 inhibits differentiation of mouse common lymphoid progenitors into CD19-positive B cells and produces B-cell developmental arrest in a mouse model, and downregulates B-cell development genes in REH cells.

    The pathogenic role of MEF2D-SS18 fusion gene in B-cell acute lymphoblastic leukemia. · Zhang M, Mao D, Zhang W · Biochem Biophys Res Commun 2018
    PMID 29408457DOI
    What this evidence cannot show

    This is MEF2D-SS18, NOT SS18-SSX and NOT any KMT2A fusion, and the localisation datum is a single immunofluorescence panel of an overexpressed construct showing a nuclear signal - it resolves no subnuclear compartment and says nothing about condensates. The B-cell arrest is mouse and REH-cell work. Wild-type SS18's own localisation in a B-lymphoid cell remains unmeasured.

    Evidence that argues against this (1)
    • Reading this as 'SS18 is nuclear in B-ALL' would be an over-read in a way that matters for this graph's differentiation-block argument: the protein imaged is a chimera whose N-terminal half is MEF2D, a sequence-specific DNA-binding transcription factor that is nuclear in its own right, so the fusion's nuclear signal is at least as attributable to MEF2D as to SS18. The SS18 half's own localisation behaviour - tyrosine-IDR-driven condensate formation - was characterised in mouse ESCs and has never been looked for in a lymphoid cell, so whether MEF2D-SS18 forms condensates in a B-ALL blast is entirely open. Kuang J et al. 2021
Connected to