Oncofusion

start / graph / gene-abi1

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

ABI1 (Abl interactor 1 / e3B1) — WAVE-complex cytoskeletal adaptor, KMT2A partner, and an ENL-binding protein

also called ABI-1 · e3B1 · SSH3BP1 · Nap1BP · AblBP4 · Abl interactor 1

ABI1 is a connector protein in the machinery that builds and steers the cell's actin skeleton. It is a rare KMT2A fusion partner, and unusually it also sticks to ENL — one of the common KMT2A partners — which could let it reach the gene-reading machinery indirectly.

The same thing, for a scientist

ABI1 (UniProt Q8IZP0, ABI1_HUMAN, 508 aa) is an Abl-interactor and a constitutive subunit of the WAVE regulatory complex. Its annotated architecture is a t-SNARE coiled-coil homology domain at aa 45-107, a region at 18-79 required for WASF1 (WAVE1) binding, three long disordered stretches (159-290, 306-375, 388-421) and a C-terminal SH3 domain at 446-505 [E1]. It has no chromatin-binding domain, no YEATS domain and no transactivation domain. Two properties bear on the KMT2A question. It carries a coiled-coil-homology region, which fits the oligomerisation template that explains EPS15 and GAS7 fusions [E3]. More distinctively, ABI1 was recovered in a yeast two-hybrid screen as a binding partner of ENL — the t(11;19) KMT2A fusion partner and a YEATS-domain SEC subunit — with the interaction confirmed by far-Western, GST pulldown and co-immunoprecipitation and mapped to an internal ENL region and a composite ABI1 motif including the SH3 domain [E2]. ABI1 is therefore the only documented physical bridge between a cytoskeletal KMT2A partner and an elongation KMT2A partner.

What it looks like
ABI1 (Abl interactor 1 / e3B1) — WAVE-complex cytoskeletal adaptor, KMT2A partner, and an ENL-binding protein508 aa
t-SNARE coiled-coil homologyt-SNARE coiled-coil homology · residues 45–107WASF1-binding regionWASF1-binding region · residues 18–79SH3SH3 · residues 446–5050200400

No three-dimensional model is available for this protein, so the map above is the whole picture. That is itself informative: AlphaFold does not model chains this long, and the parts that have been crystallised are held under the entries for the individual domains.

Go deeper
A cytoskeletal adaptor with an unexpected nuclear contact

The normal job of ABI1 is in the WAVE regulatory complex, where it links Abl-family kinases, Eps8 and Nap1 to actin nucleation. The region at aa 18-79 is required for WASF1 binding and the SH3 domain at 446-505 mediates Abl binding 1. None of that has any evident relationship to transcription.

The ENL interaction is the exception 2. It was found by an unbiased two-hybrid screen for ENL-binding proteins and confirmed three independent ways. The authors noted at the time that no general unifying property had been detected among the MLL fusion partners, and offered the ABI1-ENL contact as a novel aspect of leukemogenesis by MLL fusions. Two decades later the elongation hijack is the unifying property everyone assumes, and this interaction is the one piece of evidence that a non-elongation partner could reach it without encoding it.

What is not known

The co-immunoprecipitation was between wild-type proteins 2. Nobody has shown that the MLL-ABI1 fusion protein binds ENL, that ENL is recruited to fusion-bound chromatin, or that the SH3 domain is retained in patient fusions — the founding case used an alternatively spliced ABI1 transcript 4. This entry records a bridge that exists in principle and has never been walked across.

Facts and figures
uniprotQ8IZP0
uniprot verifiedQ8IZP0 = ABI1_HUMAN, PRIMARY gene name ABI1, 508 aa, Homo sapiens. Fetched from the UniProt REST API on 2026-09-19 and the primary gene name read back; SSH3BP1 is the only gene synonym.
length AA508
localizationcytoplasm; WAVE regulatory complex at the leading edge
familyAbl-interactor family; WAVE regulatory complex subunit
domains
WASF1-binding region18.0WASF1-binding region: 18t-SNARE coiled-coil homol…45.0t-SNARE coiled-coil homology: 45SH3446.0SH3: 446
domains — the numbers behind the chart
[
  {
    "name": "t-SNARE coiled-coil homology",
    "start": 45,
    "end": 107
  },
  {
    "name": "WASF1-binding region",
    "start": 18,
    "end": 79
  },
  {
    "name": "SH3",
    "start": 446,
    "end": 505
  }
]
essentiality
{
  "depmap_common_essential": false,
  "chronos_mean": -0.065,
  "chronos_n_lines": 1208,
  "kmt2a_selectivity_rank": 1169,
  "kmt2a_selectivity_of": 17728
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.096,
  "chronos_mean_other_leukaemia": 0.0028,
  "difference": -0.0987,
  "pan_essentiality_all_lines": -0.0649,
  "selectivity_rank": 1169,
  "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."
}
How this could bite us back
  • ABI1 is a WAVE-complex subunit with a broad cytoskeletal requirement, so any dependency signal for it in leukaemia lines is confounded by essential-function biology and is a poor basis for target selection.
  • If the ABI1-ENL bridge is real in the fusion context, it argues that partner class does not determine mechanism, which weakens the partner-stratified reasoning this graph uses to decide where a drug should work.
Still open
  1. Does the MLL-ABI1 fusion protein co-immunoprecipitate with ENL, and does it pull down the super elongation complex?
  2. Is the ABI1 SH3 domain retained in KMT2A::ABI1 fusions beyond the single sequenced index case?
  3. Does ABI1 also contact the other YEATS-domain partners MLLT1's paralogue MLLT3 or MLLT10, or is the ENL contact specific?
Every source behind this entry (4)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Human ABI1 is UniProt Q8IZP0 (ABI1_HUMAN), primary gene name ABI1, 508 aa, with a t-SNARE coiled-coil homology domain at aa 45-107, a region at aa 18-79 required for binding to WASF1, disordered regions at 159-290, 306-375 and 388-421, and an SH3 domain at aa 446-505; protein aliases include e3B1, Eps8 SH3 domain-binding protein, Nap1-binding protein and Abl-binding protein 4.

    What this evidence cannot show

    Canonical isoform; ABI1 is extensively alternatively spliced and the KMT2A fusion in the founding case used a non-canonical transcript, so these coordinates do not necessarily describe the oncoprotein. Primary gene name read back from the REST record on 2026-09-19.

    Evidence that argues against this (1)
    • The t-SNARE coiled-coil homology annotation is homology-derived and has never been shown to mediate ABI1 self-association, so it should not be equated with the experimentally validated coiled-coil oligomerisation domains of EPS15/AF1p and GAS7. So CW et al. 2003
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    ABI1 was identified as an ENL-binding protein in a yeast two-hybrid screen, with the interaction verified in vitro by far-Western blot and GST pulldown and in vivo by co-immunoprecipitation, and structure-function analysis identifying an internal region of ENL and a composite ABI1 motif including the SH3 domain as the mutual binding determinants; ENL is the MLL fusion partner in t(11;19) and ABI1 is itself fused to MLL in t(10;11)(p11.2;q23).

    PMID 10777208DOI
    What this evidence cannot show

    Interaction of the wild-type proteins. No experiment used the MLL-ABI1 or MLL-ENL fusion proteins, and no chromatin or transcription readout was performed. Yeast two-hybrid primary hits require the orthogonal confirmations the authors supplied, which they did.

    Evidence that argues against this (1)
    • Whether this interaction survives into the oncoprotein is unknown: the founding MLL-ABI1 fusion used an alternatively spliced ABI1 transcript, so the SH3-containing composite motif may not be present in patient fusion proteins. Taki T et al. 1998
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The coiled-coil oligomerisation domains of cytoplasmic KMT2A partners AF1p/EPS15 and GAS7 are necessary and sufficient for leukemogenic transformation, and synthetic dimerisation modules substituted for them reproduce Hox activation and enhanced progenitor self-renewal.

    PMID 12957285DOI
    What this evidence cannot show

    Establishes the template for cytoplasmic partners generally; ABI1 was not among the partners tested.

    Evidence that argues against this (1)
    • The template does not generalise to every cytoplasmic partner: a synthetic dimerisation module substituted for the SEPT6 coiled coil failed to immortalise myeloid progenitors, so a coiled-coil-homology region in ABI1 cannot be assumed to be sufficient. Ono R et al. 2005
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The MLL-ABI-1 fusion transcript in the founding AML case was formed by an alternatively spliced ABI-1 and combined the MLL AT-hook motifs and DNA methyltransferase homology region with the ABI-1 homeodomain-homologous region, polyproline stretches and SH3 domain.

    PMID 9694699
    What this evidence cannot show

    Single patient. No DOI is registered for this 1998 Blood article; the PMID was fetched from NCBI and is the only stable identifier.

    Evidence that argues against this (1)
    • A second MLL-ABI1 patient was reported three years later confirming recurrence, but that report did not re-characterise the fusion's domain content, so the domain composition of KMT2A::ABI1 rests on a single sequenced case. Shibuya N et al. 2001
Connected to