GATA3 - one of only two transcription-factor baits in a 109-bait BioID screen that recover DOT1L
also called GATA binding protein 3 · HDR · Trans-acting T-cell-specific transcription factor GATA-3
GATA3 is the master switch that makes a blood progenitor become a T cell. In a 2022 screen that tagged 109 human transcription factors, GATA3 was one of just two whose neighbourhood contained DOT1L - the enzyme whose mark this whole project turns on - which is a thread worth pulling even though the experiment was done in kidney cells.
The same thing, for a scientist
GATA3 (HGNC:4172, ENSG00000107485, UniProt P23771, 443 aa, 10p14) is the lineage-defining zinc-finger transcription factor of T-cell development and Th2 differentiation [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, **GATA3 bait / DOT1L prey**, MI-score 0.27, host HEK293 Flp-In [E1][E2]. What makes it worth an entry rather than a line in a table is the denominator: of **109 transcription-factor baits**, this graph's reading of the deposition records **only two - GATA3 and TBXT - recovering DOT1L**, against eleven recovering KMT2A [E1]. A 2-of-109 prey is a much more selective observation than an 11-of-109 prey, and DOT1L is the enzyme whose H3K79me2 output this project measures constantly. **That selectivity is the entire argument for this entry, and it is an argument about a screen, not about biology.** No orthogonal validation of GATA3-DOT1L exists, the observation is in embryonic kidney cells, and GATA3 is not natively expressed in that context at physiological levels.
What the interaction record actually is
One IntAct row, IM-28767, deposited 2022-01-14, last updated 2026-01-11. Detection MI:1314 proximity-dependent biotin identification; interaction type MI:2364 proximity; expansion MI:1060 spoke; experimental roles MI:0496 bait (GATA3) / MI:0498 prey (DOT1L); host organism taxid:9606(Homo sapiens HEK293flp-in embryonic kidney cell); curator annotation figure legend: Supplementary table 1B; MI-score 0.27 2.
BioID reports proximity within a labelling radius during hours of bait over-expression. It is not binding, and a chromatin enzyme in the same nuclear neighbourhood as a DNA-binding factor is a low bar. The value of this row is entirely comparative: 2 of 109 baits, not the row itself.
Why DOT1L-proximity to a lineage TF is a question this project should care about
DOT1L is the writer of H3K79me2, the mark that distinguishes KMT2A-fusion target genes and the basis of the pinometostat programme. How DOT1L is recruited to particular genes - and whether lineage transcription factors participate in that recruitment - is an open question this graph touches repeatedly.
The lineage angle sharpens it. GATA3 is the T-lineage master factor, and this graph already holds cell-type-early-t-precursor and a whole CALM-AF10 beat, because PICALM::MLLT10 drives T-lineage disease through the same DOT1L/H3K79 axis. A T-lineage TF whose neighbourhood contains DOT1L is therefore a plausible - and completely untested - component of why the H3K79 axis produces T-lineage as well as B-lineage and myeloid disease.
This entry states that as a question. It does not claim GATA3 recruits DOT1L. One over-expression proximity row in kidney cells cannot support that and nothing else was found.
Localization, cell type, cell state
UniProt P23771 assigns nucleus with no subnuclear compartment and, notably, no ECO evidence code on the location comment 3 - so this is an uncoded annotation, which this graph already knows is not a measurement. No subnuclear address is asserted here.
Cell type of the observation: HEK293 Flp-In embryonic kidney - a cell that does not normally express GATA3 as a lineage factor. Cell state: none specified. The physiological setting where this pair would matter - a T-lineage progenitor, or an ETP-ALL blast - has never been examined for it.
Other fusion families
Nobody has looked. GATA3 appears in this graph's fusion-nup98-kdm5a and cell-type-early-t-precursor entries, but this beat verified no GATA3-DOT1L result in any fusion family and asserts none. Rule 86: the honest value of shown_for_fusion here is 'nobody has looked in any fusion', not a bare NO used to imply a tested negative.
| hgnc ID | HGNC:4172 |
|---|---|
| symbol | GATA3 |
| ensembl gene ID | ENSG00000107485 |
| uniprot | P23771 |
| length AA | 443 |
| chromosomal location | 10p14 |
| identifier provenance | HGNC REST fetch/symbol/GATA3 and UniProtKB P23771 retrieved live 2026-09-23. |
| subnuclear location | none assigned; none asserted here |
| cell type of the observation | HEK293 Flp-In embryonic kidney - a cell that does not use GATA3 as a lineage factor |
| cell state | none specified; asynchronous over-expression culture |
| epigenetic role | None demonstrated for GATA3 itself in this record. The interest is the converse: whether a lineage transcription factor participates in DOT1L recruitment. |
| shown for fusion | NO - nobody has looked in any fusion. The row is wild-type DOT1L in HEK293 Flp-In, and the T-lineage context where it would matter (CALM-AF10 / PICALM::MLLT10, ETP-ALL) has never been examined for it. |
| evidence class | single publication, single method, spoke-expanded BioID proximity, no orthogonal validation |
prev and alias symbols
[ "HDR" ]
interaction with project baits
[
{
"bait": "DOT1L",
"bait_uniprot": "Q8TEK3",
"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,
"selectivity_in_the_screen": "2 of 109 TF baits recover DOT1L (GATA3 and TBXT), against 11 of 109 recovering KMT2A - figure taken second-hand from this graph's ana-tf-bioid-network entry, not recomputed"
}
]subcellular location uniprot
[ "Nucleus (NO evidence code - uncoded annotation)" ]
- GATA3 is required for normal T-cell development; nothing about this row nominates it as a target, and an intervention against a lineage master factor has an obvious immune-toxicity cost.
- If lineage transcription factors do participate in DOT1L recruitment, then DOT1L inhibition would have lineage-specific off-target consequences in normal T-cell development that a myeloid-focused toxicity assessment of pinometostat would not have seen.
- Does GATA3 co-occupy H3K79me2 domains in a T-lineage leukaemia - PICALM::MLLT10 or ETP-ALL - and does GATA3 loss move H3K79me2 there? That is the experiment this row is worth.
- Recompute the selectivity from the IM-28767 deposition directly: how many of the 109 baits recover DOT1L, and is the 2-of-109 figure robust to the MI-score threshold? This entry's central comparative claim is currently second-hand.
- Is DOT1L recovery by a bait predicted simply by bait abundance or nuclear residence time, in which case the selectivity is an artefact?
- Does the analogous myeloid or B-lineage master factor recover DOT1L in the same screen? If lineage factors generally do not and GATA3 does, that is interesting; if none of them was tested, the comparison is empty.
- 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, and state that the BioID analysis identifies more high-confidence interactions, highlighting the transient and dynamic nature of many of the TF interactions.
Human transcription factor protein interaction networks. · Goos H, Kinnunen M, Salokas K, et al. · Nat Commun 2022What this evidence cannot show
Abstract read; the supplementary tables that contain this pair were not opened - the pair was read from the IMEx deposition IM-28767. This is a 109-bait screen in one engineered background (HEK293 Flp-In), so every row is an over-expressed, tagged bait in embryonic kidney cells.
Evidence that argues against this (1)
- This graph's reading of the same deposition records KMT2A as prey of eleven of the 109 TF baits and DOT1L as prey of two (GATA3 and TBXT). A chromatin enzyme recovered by unrelated TF baits in one over-expression screen is as consistent with an abundant or promiscuous prey as with specific relationships. source 2026
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
IntAct serves one curated human GATA3-DOT1L row: MI:1314 proximity-dependent biotin identification, MI:2364 proximity, MI:1060 spoke expansion, GATA3 experimental role MI:0496 bait and DOT1L 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'.
IntAct molecular interaction database (EMBL-EBI), PSICQUIC tab27 record for id:P23771 AND id:Q8TEK3 2026What this evidence cannot show
Retrieved live 2026-09-23. Rule 93: PSICQUIC is a floor relative to IntAct REST. The '2 of 109 baits recover DOT1L' figure 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 the load-bearing comparative number in this entry and it is second-hand. Re-derive it before building on it.
Evidence that argues against this (1)
- Selectivity in a screen can be a detection-limit effect. DOT1L is a large, low-abundance nuclear protein, and recovery by only two baits is as consistent with being near the detection threshold - so that recovery is close to random among baits - as with genuine selectivity. Distinguishing these requires the deposition's intensity data, which this beat did not examine. source 2026
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt P23771 is 443 aa and records GATA3 as a transcriptional activator binding the enhancer of the T-cell receptor alpha and delta genes and the consensus 5'-AGATAG-3', required for the Th2 differentiation process, with subcellular location given as nucleus and NO ECO evidence code on that location comment.
What this evidence cannot show
Fetched live 2026-09-23. The absence of an evidence code on the location comment is recorded deliberately: this graph has already measured that uncoded UniProt localisation annotations are not measurements in about a tenth of its entries, and GATA3 is one of them.
Evidence that argues against this (1)
- UniProt's GATA3 location comment carries NO evidence code. This graph has already measured that uncoded UniProt localisation annotations are not measurements in about a tenth of its entries, so 'nucleus' here is a curator statement of the obvious rather than a cited experiment, and it cannot exclude a cytoplasmic or shuttling pool. source 2026
GATA3 bait / DOT1L prey, one spoke-expanded BioID row, IntAct IM-28767 - one of only two of 109 TF baits recovering DOT1L
the GATA3 bait row in this deposition
- expressed in →Early T-cell precursor (ETP) — the thymocyte that has not yet given up being myeloid, and where KMT2A-rearranged T-ALL sits
T-lineage master transcription factor; the cell type where a GATA3-DOT1L relation would matter
GATA family; GATA2 is the HSC-stage factor, GATA3 the T-lineage one
- ← co-occurs withTBXT / brachyury - the notochord master regulator, and one of only two of 102 transcription-factor baits whose neighbourhood contained DOT1L
the other of the two TF baits recovering DOT1L in IM-28767; GATA3-DOT1L p = 0.032, TBXT-DOT1L p = 0.039, the two smallest of 62