LDB1::KMT2A - the clearest statement in this literature that a sarcoma fusion keeps KMT2A exons 4 to 36 and discards exons 1 to 3
also called LDB1::KMT2A · LDB1-KMT2A
A single reported sarcoma joined the front of a gene called LDB1 to almost the whole of KMT2A, starting at KMT2A's fourth exon. That means it keeps KMT2A's enzyme end but loses the first three exons, including the one menin holds on to.
The same thing, for a scientist
One published case: a 19-year-old man with an 8 cm erector spinae spindle-cell sarcoma carrying LDB1 exons 1-10 fused to KMT2A exons 4-36, confirmed by FISH and RT-PCR, with a concurrent pathogenic BCOR frameshift (c.3203dup, p.E1069Gfs*10) and a Ki-67 of about 10%. Translated: the chimera contains KMT2A aa 1053-3969, which retains the CXXC zinc finger, PHD1-3, the bromodomain, the FYR domains and SET, and discards aa 1-1052 - the menin-binding motif (6-25), both LEDGF integrase-binding motifs (123-134, 147-152) and all three AT-hooks (169-309). It is included in this graph for one reason: its published exon range is the least ambiguous statement in the KMT2A-sarcoma literature that these fusions begin downstream of exon 1, and it is a third independent partner making the same architectural choice. LDB1 (LIM domain binding 1) is a transcriptional adaptor and a core component of the LDB1/LMO2/TAL1 complex in haematopoiesis - so a fourth distinct partner class, after a Hippo co-activator, an intermediate filament and a chromobox. The case was disease-free at 5 months, which is far too short to say anything about behaviour.
What the reported exon range encodes
LDB1 exons 1-10 :: KMT2A exons 4-36 1. On the exon map 2, KMT2A exon 4 begins at codon 1053, so:
| KMT2A retained | aa 1053-3969 - CXXC (1147-1195), PHD1-3, Bromo (1635-1765), FYRN/FYRC, SET (3829-3945) |
| KMT2A discarded | aa 1-1052 - menin-binding motif (6-25), IBM1 (123-134), IBM2 (147-152), AT-hooks 1-3 (169-309) |
So this is the VIM-arm architecture pushed two exons further: an even larger N-terminal deletion, still retaining the catalytic domain, still without a menin interface.
Third independent partner, same choice. YAP1 (internal insert of exons 4/5-6), VIM (exon 2 onward), LDB1 (exon 4 onward). Three partners, three different KMT2A start points, and exon 1 absent from all three.
Why the BCOR frameshift matters more than the case report treats it
The tumour carried a pathogenic BCOR frameshift alongside the fusion, and expressed CD99, SATB2, cyclin D1, BCL2, TLE1, pan-TRK, NKX2.2 and focal BCOR 1. BCOR-altered sarcoma is an established entity with exactly that kind of round-to-spindle immunoprofile. So there are two readings: LDB1::KMT2A is a novel driver, or it is an incidental rearrangement in a BCOR-driven tumour. The founding VIM::KMT2A case was also BCOR, WT1 and NKX2-2 positive 3, which makes the coincidence worth recording rather than dismissing - two of the small number of published KMT2A-sarcoma cases have a BCOR-flavoured phenotype.
What this implies for the index fusion, by contrast
LDB1 is the one sarcoma KMT2A partner that has a haematopoietic day job: it is the adaptor in the LDB1/LMO2/TAL1/GATA1 complex that this project's sibling literature on T-ALL and erythroid transcription already knows. So the same partner class that is leukaemogenic in one architecture appears here in the opposite polarity, in mesenchyme, with the menin interface removed. If any single case invites the question 'is it the partner or the polarity?', this is it - and with n=1 and a co-driver, it can only pose the question.
| gene5 | LDB1 |
|---|---|
| gene3 | KMT2A |
| architecture | conventional 5'-3' fusion with KMT2A as the 3' partner, from exon 4 |
| uniprot3 | Q03164 |
| length AA 3 | 3969 |
| menin binding motif in fusion | false |
| ledgf ibm in fusion | false |
| cxxc in fusion | true |
| set domain in fusion | true |
| partner class | LIM domain binding 1, a transcriptional adaptor and core component of the LDB1/LMO2/TAL1 complex - a fourth distinct partner class in this entity and the only one with an established haematopoietic role |
| published case count | 1 |
| note | included in this graph chiefly because its published exon range (KMT2A exons 4-36) is the least ambiguous statement in this literature that the sarcoma fusions begin downstream of KMT2A exon 1 |
| partner function | LDB1 is an architectural enhancer-looping factor, not merely a LIM-domain adaptor: it promotes connectivity among regulatory elements independently of CTCF, cohesin and YY1 [E4], and in T-ALL it is essential for proliferation and survival and drives MYB through a remote enhancer [E5]. |
| why this partner is cross family | It is the only partner in the KMT2A-sarcoma set with a documented role in a leukaemia. That role is as wild-type co-activator in T-ALL, not as a fusion partner -- LDB1 is absent from the KMT2A recombinome -- so the link is functional rather than structural. PI dimension 5 for this entry rests on the partner's biology in another disease, which is a weaker form of analogy than a shared rearrangement and is labelled as such. |
| the orientation problem this raises | In leukaemic KMT2A fusions the partner is 3-prime and is fused to KMT2A's TARGETING N-terminus (MBM, LEDGF IBMs, AT-hooks, CXXC). Here the partner is 5-prime and is fused to KMT2A's CATALYTIC C-terminus: LDB1 exons 1-10 to KMT2A aa 1053-3969, retaining CXXC through SET and discarding the menin and LEDGF tethers and all three AT-hooks. So if forced oligomerisation or forced connectivity is the conserved core across KMT2A partner classes, this fusion tests whether that core also operates in reverse -- a self-associating architectural factor driving the SET domain rather than the targeting module. Nobody has tested it, and no LDB1::KMT2A protein has been observed. |
diseases
[ "KMT2A-rearranged soft tissue sarcoma, spindle-cell" ]
breakpoints
[
{
"id": "bp-ldb1-ex10-kmt2a-ex4",
"gene5_exon_last": 10,
"gene3_exon_first": 4,
"frequency_pct": 100,
"population": "the single published LDB1::KMT2A case",
"gene3_residues_retained": [
1053,
3969
],
"domains_retained_3": [
"CXXC",
"PHD1",
"PHD2",
"PHD3",
"Bromodomain",
"FYRN",
"FYRC",
"SET"
],
"domains_lost_3": [
"menin-binding motif (6-25)",
"IBM1 (123-134)",
"IBM2 (147-152)",
"AT-hooks 1-3 (169-309)"
],
"phenotype_notes": "biphasic hypercellular fascicular and hypocellular fibroma-like areas; CD99+, SATB2+, cyclin D1+, BCL2+, TLE1+, pan-TRK+, NKX2.2+, focal BCOR; Ki-67 ~10%"
}
]co mutations
[ "BCOR (c.3203dup, p.E1069Gfs*10, pathogenic frameshift)" ]
- This case retains the SET domain and a BCOR frameshift at once, so a KMT2A-catalytic-inhibitor rationale here would be competing with an independent chromatin driver - a single-agent proposal built on the fusion alone would be naive.
- With one case and 5 months of follow-up, any claim about this arrangement's aggressiveness - in either direction - is unsupported, and an indolent-looking single case is the kind of datum that quietly lowers a clinician's guard.
- Is LDB1::KMT2A a driver or a passenger beside the BCOR frameshift? With one case the only route is finding a second, and this literature grows by roughly one case report a year.
- LDB1 is a haematopoietic transcriptional adaptor. Would LDB1::KMT2A in the 5'-KMT2A polarity be leukaemogenic? Nobody has reported an LDB1 KMT2A fusion in leukaemia, which is itself worth checking properly.
- Two of the published KMT2A-sarcoma cases have a BCOR-flavoured immunoprofile with different fusion partners. Coincidence of differential diagnosis, or a shared transcriptional state?
- Do LDB1 exons 1-10 retain the self-association surface the looping function depends on? The analogy to forced oligomerisation stands or falls on that, and this graph has not verified the residue boundaries against the retained exons.
- Does LDB1::KMT2A drive a MYB programme, given that wild-type LDB1 controls MYB through a remote enhancer in T-ALL? That is a cheap RNA-seq question on the one archival case.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In a 19-year-old man with an 8 cm left erector spinae spindle-cell sarcoma of biphasic hypercellular and fibroma-like architecture, next-generation sequencing identified a novel LDB1::KMT2A fusion involving exons 1-10 of LDB1 and exons 4-36 of KMT2A, verified by FISH and RT-PCR, with a concurrent pathogenic BCOR frameshift (c.3203dup, p.E1069Gfs*10); the tumour was CD99, SATB2, cyclin D1, BCL2, TLE1, pan-TRK and NKX2.2 positive with focal BCOR, Ki-67 about 10%, and the patient was disease-free 5 months after wide excision.
LDB1::KMT2A Fusion in a Spindle-Cell Sarcoma: A Case Report. · Wang Z, Zhang Y, Zhu M, et al. · Genes Chromosomes Cancer 2025What this evidence cannot show
A single case report read at abstract level. Five months of follow-up says nothing about natural history. The concurrent pathogenic BCOR frameshift means the fusion is not the only driver, and the immunoprofile overlaps BCOR-altered sarcoma, synovial sarcoma (TLE1) and Ewing-family tumours (CD99, NKX2.2). No functional work.
Evidence that argues against this (1)
- The same authors note that the entity's clinical behaviour ranges from surgical cure to metastasis, and this patient was disease-free at 5 months after surgery alone - so the case cannot distinguish a novel driver from an incidental rearrangement in a BCOR-driven tumour, and the paper does not claim to. Nakamura H et al. 2023
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Computed here (2026-09-21): KMT2A exon 4 of transcript ENST00000534358 begins at codon 1053, so an LDB1(ex1-10)::KMT2A(ex4-36) transcript encodes KMT2A aa 1053-3969 - retaining the CXXC zinc finger, PHD1-3, bromodomain, FYR domains and SET - and discards aa 1-1052, which contains the menin-binding motif, both LEDGF integrase-binding motifs and all three AT-hooks.
Ensembl REST (ENST00000534358) and UniProtKB (Q03164) records, KMT2A exon boundaries intersected with the feature table · Ensembl (EMBL-EBI); UniProt Consortium · Ensembl REST / UniProtKB 2026What this evidence cannot show
A coordinate derivation from a reported exon range in a single abstract; the precise junction sequence and frame were not published in the abstract and the full text was not obtained. No protein has been observed.
Evidence that argues against this (1)
- The abstract gives an exon range but no transcript accession, so the mapping assumes the 36-exon transcript the rest of this literature uses. If the authors numbered from a different transcript the residue boundary shifts, though the qualitative conclusion - exon 1 absent - would survive any of the transcripts in current use. Massoth LR et al. 2020
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The founding VIM::KMT2A sarcoma case was also immunopositive for BCOR, Wilms tumour 1 and NKX2-2, so a BCOR-flavoured immunoprofile has now appeared in two of the small number of published KMT2A-rearranged sarcoma cases with different partners.
KMT2A (MLL) fusions in aggressive sarcomas in young adults. · Yoshida A, Arai Y, Tanzawa Y, et al. · Histopathology 2019What this evidence cannot show
Two cases, different partners, different reports, and BCOR immunopositivity is not the same as a BCOR alteration - only the LDB1 case has a sequenced BCOR frameshift. Read from the abstract; the founding paper's full text is unreachable through all five routes (no PMCID; Wiley DOI returns HTTP 403). This is a pattern worth watching, not a finding.
Evidence that argues against this (1)
- NKX2-2, CD99 and BCOR positivity are shared by several small round cell sarcoma classes, so a shared immunoprofile across two KMT2A-sarcoma cases could equally reflect the differential diagnosis the pathologists were working through rather than shared biology. Almohsen SS et al. 2023
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
LDB1 is a direct architectural looping factor rather than a passive adaptor: acute degradation in a murine erythroid model shows it 'directly and broadly promotes connectivity among regulatory elements', and most LDB1-mediated contacts -- including some spanning hundreds of kilobases -- form in the ABSENCE of CTCF, cohesin or YY1 across multiple degron systems, with an engineered LDB1-driven chromatin loop being cohesin independent.
LDB1 establishes multi-enhancer networks to regulate gene expression · Aboreden NG, Lam JC, Goel VY, et al. · Molecular Cell 2025What this evidence cannot show
Murine erythroid cells, which is LDB1's best-characterised setting and not a mesenchymal one, and the mechanism is established for wild-type LDB1 rather than for any LDB1 fusion. Whether LDB1 exons 1-10 as retained in this chimera preserve the self-association surface that the looping function requires has not been tested, and this graph has not verified the residue boundaries of that surface against the retained exons. Title, journal, first author and DOI verified live by efetch on 2026-09-21.
Evidence that argues against this (1)
- A looping function attributed to the 5-prime partner is the least testable claim available here, because no LDB1::KMT2A fusion protein has been observed and the single case has no chromatin data of any kind. The architectural role is also exactly the sort of function that a truncated partner is most likely to lose, since it depends on an intact self-association interface rather than on a short peptide motif.
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
LDB1 has a documented oncogenic role in a leukaemia, which makes this the one sarcoma partner with cross-family biology: in T-ALL cell lines LDB1 is a critical transcription co-activator essential for proliferation and survival, cooperating with the master factors ERG, ETV6 and IRF1, and it 'controls MYB through remote enhancer modulation'.
Enhancer looping protein LDB1 modulates MYB expression in T-ALL cell lines in vitro by cooperating with master transcription factors · Li Y, Zhang Y, Yu Y, et al. · Journal of Experimental & Clinical Cancer Research 2024What this evidence cannot show
T-ALL CELL LINES, and the paper's own title says 'in vitro' -- the abstract mentions in vivo experiments for the proliferation/survival claim but the headline mechanism is cell line work. LDB1 here is a wild-type co-activator that is required in T-ALL, NOT a fusion partner: LDB1 is not a member of the KMT2A leukaemia recombinome, and a PubMed query pairing LDB1 with KMT2A/MLL and leukaemia/fusion returns nothing. So the cross-family link is at the level of the protein's function, not of a shared rearrangement. Title, journal, first author and DOI verified live by efetch on 2026-09-21.
Evidence that argues against this (1)
- This cuts against reading the sarcoma partner set as purely mesenchymal biology, but it equally undermines any claim that LDB1 was selected for a KMT2A-specific reason. A protein that is broadly required for enhancer connectivity in several lineages is exactly the kind of highly expressed, widely used factor that turns up as a fusion 5-prime partner by proximity and promoter strength rather than by functional selection -- which is the source authors' own 'unclear significance' verdict on the non-recurrent partners, and it applies here too.
- part of →KMT2A / MLL1 — the gene that breaks
KMT2A exons 4-36 = aa 1053-3969; exons 1-3 discarded
single published case, spindle-cell sarcoma in a 19-year-old man
the least ambiguous published exon range showing a sarcoma fusion starts downstream of KMT2A exon 1
exon 4 = codon 1053, which is how the retained domains were assigned
single-case partner in the census
supplies LDB1 aa 1-335 (self-association region); the LIM interaction domain in exon 11 is lost
- analogous to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
INVERTED ORIENTATION, AND THE PARTNER IS AN ARCHITECTURAL FACTOR RATHER THAN AN ELONGATION ONE: the index fusion joins KMT2A's targeting N-terminus (MBM, LEDGF IBMs, AT-hooks, CXXC) to an elongation partner. LDB1::KMT2A joins a self-associating enhancer-looping factor to KMT2A aa 1053-3969 -- CXXC through SET -- discarding the menin and LEDGF tethers and all three AT-hooks. Same gene, opposite halves, opposite partner class.
TESTS THE CONSERVED CORE IN REVERSE: if forced oligomerisation is what KMT2A partner classes share, LDB1 is a strong instance of the partner side -- it promotes connectivity among regulatory elements independently of CTCF, cohesin and YY1, and an engineered LDB1-driven loop is cohesin independent. But here it drives KMT2A's CATALYTIC C-terminus rather than its targeting module, so it asks whether the core operates when the oligomeriser is fused to SET rather than to CXXC. Untested, and no LDB1::KMT2A protein has ever been observed.
- analogous to →Mechanism class 5 — enhancer hijacking (inv(3) GATA2/EVI1 — and DUX4-r, which only half belongs here)
THE PARTNER IS LITERALLY AN ENHANCER-CONNECTIVITY PROTEIN: LDB1 directly promotes contacts among regulatory elements over hundreds of kilobases without CTCF, cohesin or YY1, so a fusion that retains its N-terminus is the most mechanistically direct enhancer-hijacking candidate in the KMT2A-sarcoma set. Stated as an analogy and not a finding -- no chromatin data of any kind exists for the single published case.
- analogous to →VIM::KMT2A - the one KMT2A rearrangement that keeps the SET domain and still throws away the menin-binding exon
THE THIRD ARM MAKING THE SAME ARCHITECTURAL CHOICE, AND THE MOST EXTREME OF THEM: both are conventional KMT2A-3-prime sarcoma fusions retaining SET and discarding the menin tether, but VIM::KMT2A keeps KMT2A from aa 145 (retaining IBM2 and all three AT-hooks) while LDB1::KMT2A keeps it only from aa 1053, losing the AT-hooks as well. The pair grades how much of the targeting module a sarcoma arm can discard.
CROSS-FAMILY BIOLOGY THROUGH THE PARTNER, NOT THROUGH A SHARED REARRANGEMENT: LDB1 is the only partner in the KMT2A-sarcoma set with a documented oncogenic role in a leukaemia -- in T-ALL it is essential for proliferation and survival, cooperates with ERG, ETV6 and IRF1, and controls MYB through remote enhancer modulation. LDB1 is NOT in the KMT2A recombinome, so this is a functional link and a weaker form of analogy than a shared fusion, which is how it is labelled.