Oncofusion

start / graph / fusion-kmt2a-septin5

Fusionmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q-nonelongation-partners

KMT2A::SEPTIN5 (MLL-SEPT5 / MLL-hCDCrel-1), t(11;22)(q23;q11.2)

also called MLL-SEPT5 · MLL-CDCREL · MLL-hCDCrel-1 · MLL-PNUTL1 · KMT2A-SEPT5 · t(11;22)(q23;q11.2)

The first septin ever found fused to KMT2A, identified in infant twins with the same leukaemia — strong evidence that the break happened once, before birth, in a shared blood cell. The partner is a cell-division septin that also happens to sit in the stretch of chromosome 22 deleted in DiGeorge syndrome.

The same thing, for a scientist

KMT2A::SEPTIN5 was the first KMT2A-septin fusion described, cloned by panhandle variant PCR from the t(11;22)(q23;q11.2) of infant twins with AML [E1]. The concordant rearrangement was detectable in the second twin before clinical disease, and its intensity relative to the normal fragment showed it was not constitutional — the canonical demonstration of a shared in utero clonal origin for a KMT2A fusion. The MLL genomic breakpoint was in intron 7 and MLL exon 7 was joined in frame to hCDCrel exon 3 [E1]. It is the third most frequent septin partner in the DCAL recombinome, n=8 of 3401 (2 infant AML, 1 paediatric ALL, 3 paediatric AML, 2 adult AML) [E2] — and notably it is the only septin partner with a documented ALL case. Clinically it remains a case-report entity: de novo adult AML, de novo paediatric AML, infant AML and MDS have all been reported since [E3][E4][E5]. SEPTIN5 (UniProt Q99719, 369 aa) is the shortest KMT2A-partner septin, with a G domain at 41-314 and a short coiled coil at 338-369 [E6]. There is no functional model of this fusion.

Go deeper
The twins

Both infant twins had two identical MLL gene rearrangements by Southern blot, indicating a chromosomal translocation; the rearrangement was already detectable in the second twin before signs of clinical disease, and the band intensity relative to the normal fragment showed it was not constitutional 1. This is one of the foundational observations of the infant-leukaemia field — a fusion arising in a single fetal cell and shared through the placental circulation — and it happens to have been made on a septin partner rather than on AFF1.

Both MLL and hCDCrel contained homologous CT, TTTGTG and GAA sequences within a few base pairs of their breakpoints 1.

Frequency and the one ALL case

Eight patients in 3401: 2 infant AML, 1 paediatric ALL, 3 paediatric AML, 2 adult AML 2. The paediatric ALL case is notable because SEPTIN6 (n=19) and SEPTIN9 (n=14) have zero lymphoid cases between them; with SEPTIN11's two ALL cases it is one of only three lymphoid septin-fusion patients in the entire series. The septin class is overwhelmingly but not absolutely myeloid.

Where the partner sits

SEPTIN5 maps to 22q11.2 within the region commonly deleted in DiGeorge and velocardiofacial syndromes 1. That is an incidental genomic coincidence rather than a mechanism, but it means constitutional 22q11.2 deletion carriers have one fewer SEPTIN5 allele, and it is the reason the gene was on the map before the leukaemia work. No link between 22q11.2 deletion syndrome and KMT2A-rearranged leukaemia has been established and this entry asserts none.

Facts and figures
gene5KMT2A
gene3SEPTIN5
cytobandt(11;22)(q23;q11.2)
uniprot5Q03164
uniprot3Q99719
uniprot3 verifiedQ99719 = SEPT5_HUMAN, PRIMARY gene name SEPTIN5, 369 aa, Homo sapiens (UniProt REST, checked 2026-09-19).
prognosisNo estimate exists. Case reports only.
PDB noteNo structure of SEPTIN5 alone, of the fusion, or of any KMT2A-SEPTIN5 interface was identified. The septin structures available (2QAG, 2QNR, 6UPA, 5CYO) cover SEPT2, SEPT6, SEPT7 and SEPT9 only.
mechanism class noteAssigned by analogy only. There is no functional model of KMT2A::SEPTIN5 — no transformation assay, no mouse, no chromatin readout.
druggability classunexplored
stage of hematopoiesisfetal haematopoietic cell for the index twins (in utero origin demonstrated); myeloid progenitor otherwise
age peakinfancy to childhood, with adult cases
frequency
paediatric_all_n1.000paediatric_all_n: 1infant_aml_n2.000infant_aml_n: 2adult_aml_n2.000adult_aml_n: 2paediatric_aml_n3.000paediatric_aml_n: 3dcal_recombinome_n8.000dcal_recombinome_n: 8dcal_recombinome_total3,401dcal_recombinome_total: 3401
diseases
[
  "infant AML",
  "paediatric AML",
  "adult AML",
  "paediatric ALL (one case)",
  "MDS"
]
frequency — the numbers behind the chart
{
  "dcal_recombinome_n": 8,
  "dcal_recombinome_total": 3401,
  "infant_aml_n": 2,
  "paediatric_all_n": 1,
  "paediatric_aml_n": 3,
  "adult_aml_n": 2
}
breakpoints
[
  {
    "id": "bp-kmt2a-intron-7-septin5-twins",
    "gene5_exon_last": 7,
    "gene3_exon_first": 3,
    "population": "the index infant-twin pair; a single clonal event, not a frequency",
    "domains_retained_5": [
      "MBM1",
      "MBM2",
      "AT-hook 1",
      "AT-hook 2",
      "AT-hook 3",
      "SNL1/2"
    ],
    "domains_lost_5": [
      "PHD1",
      "PHD2",
      "PHD3",
      "Bromodomain",
      "PHD4",
      "FYRN",
      "TAD",
      "FYRC",
      "WIN",
      "SET"
    ],
    "domains_retained_3": [
      "Septin-type G domain (aa 41-314)",
      "G1/P-loop motif (aa 51-58)",
      "G3 motif (aa 108-111)",
      "G4 motif (aa 189-192)",
      "coiled coil (aa 338-369)"
    ],
    "domains_lost_3": [
      "SEPTIN5 exons 1-2"
    ],
    "phenotype_notes": "CXXC is omitted from both arrays: the exon-7 junction lies near the CXXC-encoding interval and no primary source resolves its fate for this partner."
  }
]
PDB IDS
[]
mechanism class
[
  "forced-oligomerisation"
]
drugs
[]
key interactors
[
  "gene-men1"
]
trials
[]
How this could bite us back
  • The twin observation is used across this field as the proof that KMT2A fusions arise in utero. If that argument is to be carried over to menin-inhibitor timing or to prenatal-origin reasoning, note that it rests on a partner nobody has ever shown to transform anything.
  • SEPTIN5 sits inside the 22q11.2 DiGeorge deletion interval. Any therapeutic strategy touching SEPTIN5 dosage would land on a gene already hemizygous in a sizeable constitutional-deletion population, in whom the safety margin is unmeasured.
Still open
  1. Does KMT2A::SEPTIN5 transform at all in a model system? No assay has ever been run on this fusion, which is a striking gap for the first septin partner ever described.
  2. The SEPTIN5 coiled coil is 32 residues against SEPTIN6's 89. If coiled-coil length matters for fusion oligomerisation, SEPTIN5 and SEPTIN6 fusions should differ measurably.
  3. Is the single paediatric ALL case real lymphoid disease or a mixed-phenotype presentation classified as ALL?
Every source behind this entry (7)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In AML of infant twins, Southern blot showed two identical MLL rearrangements in both, detectable in the second twin before clinical disease and not constitutional; panhandle variant PCR cloned a t(11;22)(q23;q11.2) breakpoint with the MLL genomic breakpoint in intron 7 and an in-frame fusion of MLL exon 7 to hCDCrel exon 3, with homologous CT, TTTGTG and GAA sequences within a few base pairs of both breakpoints; hCDCrel maps to the region commonly deleted in DiGeorge and velocardiofacial syndromes.

    PMID 9600980DOI
    What this evidence cannot show

    Two patients, who are monozygotic twins and therefore one clonal event. The sequence motifs at the breakpoints are an observation on two junctions and are not established as a recombination mechanism.

    Evidence that argues against this (1)
    • The intron-7 MLL breakpoint of this index pair is not where SEPTIN5 fusions generally fall: the 3401-patient recombinome places the great majority of KMT2A breaks in the major breakpoint cluster region around introns 9-11, and gives no evidence that SEPTIN5 prefers intron 7. Meyer C et al. 2023
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In 3401 genomically characterised KMT2A-rearranged patients, SEPTIN5 was the 16th most frequent partner with n=8: 2 infant AML, 1 paediatric ALL, 3 paediatric AML and 2 adult AML.

    The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al. · Leukemia 2023
    PMID 37019990DOI
    What this evidence cannot show

    Referral cohort. Eight patients is a small denominator and the single ALL case could be a classification artefact in a biphenotypic presentation.

    Evidence that argues against this (1)
    • Case-report accumulation suggests broader phenotypic range than the recombinome classification captures, including myelodysplastic syndrome, which the recombinome's ALL/AML/other scheme does not resolve. Zou D et al. 2021
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    KMT2A-SEPT5 fusion occurs in de novo AML with t(11;22)(q23;q11.2) and is rare, with few reported cases; of 135 identified KMT2A rearrangements and 94 molecularly characterised partner genes, 35 occur recurrently but only 9 specific fusions account for more than 90% of cases.

    Acute myeloid leukemia with KMT2A-SEPT5 translocation: A case report and review of the literature. · Elzamly S, Chavali S, Tonk V, et al. · SAGE Open Med Case Rep 2018
    PMID 29326822DOI
    What this evidence cannot show

    A case report with a literature review; the partner-gene counts quoted are from the 2018-era recombinome and have since been superseded (107 in-frame fusions by 2023).

    Evidence that argues against this (1)
    • The partner census has moved: the 2023 recombinome reports 107 in-frame KMT2A gene fusions, 37 recurrent and 63 seen only once, so figures quoted from 2018 reviews understate the tail. Meyer C et al. 2023
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    MLL-SEPT5 fusion transcript has been reported in infant acute myeloid leukaemia with t(11;22)(q23;q11).

    MLL-SEPT5 fusion transcript in infant acute myeloid leukemia with t(11;22)(q23;q11). · Launay E, Henry C, Meyer C, et al. · Leuk Lymphoma 2014
    PMID 23725386DOI
    What this evidence cannot show

    Single case report, title-and-abstract level. Cited for the existence of the infant AML presentation only.

    Evidence that argues against this (1)
    • Single case reports cannot establish an age distribution; the systematic series puts only 2 of 8 SEPTIN5 patients in the infant class, with the remainder paediatric and adult. Meyer C et al. 2023
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    MLL-SEPT5 fusion transcript has been reported in a myelodysplastic syndrome patient with t(11;22)(q23;q11).

    MLL-SEPT5 Fusion Transcript in Myelodysplastic Syndrome Patient With t(11;22)(q23;q11). · Zou D, Chen Y, Wu N, et al. · Front Med (Lausanne) 2021
    PMID 35004749DOI
    What this evidence cannot show

    Single case report, title-and-abstract level. Cited for phenotypic range only.

    Evidence that argues against this (1)
    • MDS presentations of KMT2A-septin fusions are reported for SEPTIN9 as well, so this is a property of the class rather than of SEPTIN5, and none of these reports includes a functional demonstration that the fusion drives the dysplastic phenotype. Kreuziger LM et al. 2007
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Human septin-5 is UniProt Q99719 (SEPT5_HUMAN), primary gene name SEPTIN5, 369 aa, protein aliases CDCrel-1 and peanut-like protein 1, with a septin-type G domain at aa 41-314 (G1/P-loop 51-58, G3 108-111, G4 189-192) and a coiled coil at aa 338-369.

    What this evidence cannot show

    Canonical isoform. Resolved by reading back the PRIMARY gene name; PNUTL1 and the retired SEPT5 are synonyms on the record.

    Evidence that argues against this (1)
    • The annotated coiled coil is only 32 residues, far shorter than SEPTIN6's (89 aa) or SEPTIN11's (96 aa), so if the coiled coil is the transforming element the KMT2A-partner septins are not equivalent — and the coiled coil is in any case dispensable for native septin filament assembly. Sirajuddin M et al. 2007
  7. E7weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset

    In a genome-wide DepMap CRISPR (Chronos) scan, SEPTIN5 had a mean gene effect of -0.116 across 1090 lines, a KMT2A-rearranged group mean of -0.154 versus -0.088 in other leukaemia lines, and ranked 2,433 of 17,728 for KMT2A selectivity (13.7th percentile) — the most KMT2A-leaning of the KMT2A-partner septins, but nowhere near a dependency.

    What this evidence cannot show

    A difference of -0.066 in Chronos units is well inside noise for 11 lines. No KMT2A-SEPTIN5 line exists in DepMap. This is reported to close the question, not to open one.

    Evidence that argues against this (1)
Connected to