Oncofusion

start / graph / gene-elf2

Geneweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q132-interactome-tail-promotion

ELF2 / NERF - an ETS factor that works with RUNX1 on a B-cell kinase promoter, and one of eleven TF baits recovering KMT2A

also called E74 like ETS transcription factor 2 · NERF · NERF-1A · NERF-1B · NERF-2 · EU32 · ETS-related transcription factor Elf-2

ELF2 is a transcription factor that switches on genes B cells need, working together with RUNX1 on the promoter of a B-cell signalling kinase. It is one of eleven transcription factors whose neighbourhood contained KMT2A in a 2022 screen - and unlike most of those eleven, its day job is in the B lineage this project's fusion produces.

The same thing, for a scientist

ELF2 (HGNC:3317, ENSG00000109381, UniProt Q15723, 593 aa, 4q31.1), also called NERF, is an ETS-family transcription factor. UniProt records isoform 1 transcriptionally activating the **LYN** and **BLK** promoters and acting **synergistically with RUNX1 to transactivate the BLK promoter**, while isoform 2 may repress RUNX1-mediated transactivation [E3]. It enters this project's core-complex interactome on one curated row: IntAct IM-28767 from Goos et al. 2022, **MI:1314 BioID**, MI:2364 proximity, MI:1060 spoke, **ELF2 bait / KMT2A prey**, MI-score 0.27, host HEK293 Flp-In [E1][E2] - one of eleven TF baits in that screen recovering KMT2A [E1]. **On the interaction evidence alone this is indistinguishable from the other ten.** What separates it is lineage: LYN and BLK are B-cell receptor-proximal kinases, RUNX1 is the partner it works through, and KMT2A::AFF1 is a B-lineage disease whose blasts are arrested at a pro-B stage. An ETS factor operating on B-lineage kinase promoters with RUNX1 is a more specific reason to look than proximity in a kidney cell provides on its own.

Go deeper
What the interaction record actually is

One IntAct row, IM-28767. Detection MI:1314 proximity-dependent biotin identification; interaction type MI:2364 proximity; expansion MI:1060 spoke; experimental roles MI:0496 bait (ELF2) / MI:0498 prey (KMT2A); host taxid:9606(Homo sapiens HEK293flp-in embryonic kidney cell); curator annotation figure legend: Supplementary table 1B; MI-score 0.27 2.

Proximity, over-expression, embryonic kidney. Not binding, not a leukaemia cell, not the fusion. This row is the reason the entry exists and it is not the reason to believe anything.

The ETS block, and why writing one member is right

Of the eleven TF baits recovering KMT2A in this screen, three are ETS-family - ELF1, ELF2 and FEV - and two are KLF-family, KLF3 and KLF12 1. Family clustering in a prey list is ambiguous: it can mean the family shares a real relationship, or it can mean family members share an over-expression behaviour, a nuclear abundance, or a DNA-binding site distribution that puts them all in the same neighbourhood.

The standing granularity rule says one entry per thing, so this entry is ELF2 only. gene-elf1 and the others remain unwritten and are named here so the next beat can find them rather than rediscover them. The family-level question - are the ETS rows one observation or three - is an analysis, not an entry, and it is listed as an open question below.

Lineage: the one axis on which ELF2 is not interchangeable

UniProt's function comment for ELF2 is unusually specific for an ETS factor: isoform 1 transactivates LYN and BLK, and acts synergistically with RUNX1 on BLK; isoform 2 may repress RUNX1-mediated transactivation 3. Both LYN and BLK are Src-family kinases of the B-cell receptor pathway, and RUNX1 is a transcription factor this graph already carries an entry for and which is itself a recurrent fusion partner in paediatric leukaemia.

Two isoforms with opposite signs on the same RUNX1 axis is the part worth recording. The BioID deposition used one construct per bait, so the row cannot say which ELF2 isoform was near KMT2A, and the two isoforms would imply opposite predictions.

Localization, cell type, other fusion families

UniProt assigns nucleus, with no evidence code on the location comment 3 - an uncoded annotation, which this graph already knows is not a measurement. No subnuclear compartment is assigned and none is asserted.

Cell type of the observation: HEK293 Flp-In embryonic kidney. Nothing B-lineage, nothing leukaemic, nothing KMT2A-rearranged. Other fusion families: nobody has looked. The corpus mentions ELF2 only inside reciprocal-allele entries, which concern a different question entirely.

Facts and figures
hgnc IDHGNC:3317
symbolELF2
ensembl gene IDENSG00000109381
uniprotQ15723
length AA593
chromosomal location4q31.1
identifier provenanceHGNC REST fetch/symbol/ELF2 and UniProtKB Q15723 retrieved live 2026-09-23. Every alias was checked against the corpus id set; none resolves to an existing entry, so no duplicate was created.
subnuclear locationnone assigned; none asserted here
cell type of the observationHEK293 Flp-In embryonic kidney - not B-lineage, not leukaemic, not KMT2A-rearranged
epigenetic roleNone demonstrated. ELF2 is a sequence-specific transcription factor with no recorded chromatin-modifying or chromatin-reading activity.
shown for fusionNO - nobody has looked in any fusion. The row is wild-type KMT2A in HEK293 Flp-In.
evidence classsingle publication, single method, spoke-expanded BioID proximity, no orthogonal validation
prev and alias symbols
[
  "EU32",
  "NERF",
  "NERF-2",
  "NERF-1A",
  "NERF-1B"
]
interaction with project baits
[
  {
    "bait": "KMT2A",
    "bait_uniprot": "Q03164",
    "partner_experimental_role": "bait",
    "scaffold_experimental_role": "prey",
    "method": "MI:1314 proximity-dependent biotin identification",
    "interaction_type": "MI:2364 proximity",
    "expansion": "MI:1060 spoke",
    "mi_score": 0.27,
    "imex": "IM-28767",
    "pmid": "35140242",
    "host_cell": "HEK293 Flp-In embryonic kidney",
    "n_rows": 1,
    "n_publications": 1
  }
]
subcellular location uniprot
[
  "Nucleus (NO evidence code - uncoded annotation)"
]
sibling preys in the same screen still unwritten
[
  "ELF1",
  "FEV",
  "KLF3",
  "KLF12",
  "EN1",
  "NHLH1",
  "NFIX",
  "NFYC"
]
How this could bite us back
  • LYN and BLK are B-cell receptor-proximal kinases and are targeted by existing inhibitors; a rationale built on one over-expression proximity row would be a bad reason to reach for one, and this entry should not be cited as support.
  • Any ETS-family intervention faces immediate redundancy - the family is large, the binding sites are shared, and this screen itself recovered three ETS factors in the same prey list.
Still open
  1. Are the three ETS rows (ELF1, ELF2, FEV) and the two KLF rows (KLF3, KLF12) in this screen one observation or five? A single analysis over the IM-28767 deposition - regressing prey recovery on bait abundance, nuclear residence and DNA-binding-site overlap - would decide whether these eleven rows carry information, and would justify writing or declining eight more partners at once.
  2. Which ELF2 isoform was the bait? Isoform 1 synergises with RUNX1 and isoform 2 may repress it, so the two imply opposite predictions and the deposition used one construct per bait.
  3. Is ELF2 expressed in KMT2A::AFF1 infant ALL blasts, and does it occupy the LYN or BLK promoters there? The B-lineage kinase link is the only non-generic reason to care about this factor and it has never been checked in the index disease.
  4. Does KMT2A::AFF1 alter LYN or BLK expression? If it does, an ELF2-RUNX1 axis on those promoters becomes a testable route rather than a coincidence.
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Goos et al. identified 6,703 BioID and 1,536 AP-MS protein-protein interactions for 109 different human transcription factors; in the resulting IntAct deposition KMT2A is recovered as prey by eleven of those TF baits - EN1, NHLH1, ELF1, ELF2, FEV, KLF12, KLF3, YY1, MYC, NFIX and NFYC.

    Human transcription factor protein interaction networks. · Goos H, Kinnunen M, Salokas K, et al. · Nat Commun 2022
    PMID 35140242DOI
    What this evidence cannot show

    Abstract read for the screen description; the eleven-bait enumeration is taken from this graph's ana-tf-bioid-network-reaches-kmt2a-and-af10 entry and was NOT independently recomputed from the deposition by this beat. It is second-hand and should be re-derived before being used as a denominator. Every row is an over-expressed tagged bait in HEK293 Flp-In embryonic kidney cells.

    Evidence that argues against this (1)
    • Eleven of 109 unrelated transcription factors recovering the same large nuclear protein is at least as consistent with KMT2A being an abundant or promiscuous prey as with eleven specific relationships, and this screen cannot distinguish them. source 2026
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    IntAct serves one curated human ELF2-KMT2A row: MI:1314 proximity-dependent biotin identification, MI:2364 proximity, MI:1060 spoke expansion, ELF2 experimental role MI:0496 bait and KMT2A MI:0498 prey, MI-score 0.27, host taxid:9606 Homo sapiens HEK293flp-in embryonic kidney cell, IMEx IM-28767, publication PMID 35140242, curator annotation 'Supplementary table 1B'.

    What this evidence cannot show

    Retrieved live 2026-09-23. Rule 93: PSICQUIC is a floor relative to IntAct REST. MI-score 0.27 is the deposition-wide value and carries no pair-specific information.

    Evidence that argues against this (1)
    • A single row is what this endpoint returns, not what exists. Rule 93 established that IntAct's REST endpoint returns roughly 1.6x the rows PSICQUIC does for these baits and reaches processed-chain identifiers PSICQUIC cannot, so a REST query on Q15723/Q03164 could return more rows, a different method, or an experimental role that contradicts the one quoted here. That query was not run for this pair. source 2026
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    UniProt Q15723 is 593 aa and records that ELF2 isoform 1 transcriptionally activates the LYN and BLK promoters and acts synergistically with RUNX1 to transactivate the BLK promoter, while isoform 2 may function in repression of RUNX1-mediated transactivation; subcellular location is given as nucleus with no evidence code.

    What this evidence cannot show

    Fetched live 2026-09-23. The function comments carry no evidence codes in the fields retrieved and the underlying primary papers were NOT opened by this beat, so this is a database assertion rather than a read experiment. The isoform-2 repression statement is hedged in UniProt's own wording ('may function in'), and that hedge is preserved here. The nucleus call is uncoded - not a measurement.

    Evidence that argues against this (1)
    • A promoter transactivation result in a reporter assay does not establish that ELF2 controls LYN or BLK in a primary B-lineage cell, and this entry has not checked whether any such in vivo evidence exists. source 2026
Connected to