ProteinstrongQ2b-atlas2

PBX1 — domain architecture and residue coordinates

also called PBX1 · pre-B-cell leukemia transcription factor 1 · PRL · P40424

PBX1 is a DNA-gripping protein that works as a partner, holding HOX proteins in place so they land on the right genes. In one childhood leukaemia its own front end is cut away and replaced by a loud activation module from TCF3, turning a positioning device into a permanent amplifier.

The same thing, for a scientist

PBX1 (UniProt P40424, 430 aa) is a TALE-class homeodomain protein comprising the PBC homology domain (38-232, subdivided into PBC-A at 45-124 and PBC-B at 127-232) and a TALE-type homeodomain at 233-295. UniProt annotates the TCF3::PBX1 breakpoint between residues **88 and 89**, so the fusion retains PBX1 89-430: PBC-A is cut roughly in half, PBC-B and the homeodomain survive intact. There is a documented and important disagreement about the DNA-binding domain's extent: UniProt calls it 233-295, whereas Piper and colleagues showed crystallographically that 'the Pbx1 DNA-binding domain is larger than the canonical homeodomain, containing an additional alpha helix', and their structure 1B72 models PBX1 to residue **319**.

What it looks like

PBX1430 aa

loading AF-P40424-F1…

AlphaFold AF-P40424-F1 ↗

colourshape

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

very high (90+)confident (70–90)low (50–70)very low (<50)

I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper

A fourth helix that UniProt does not draw

This is the clearest UniProt-versus-structure conflict in the comparator set, and the structure paper states it explicitly.

Piper and colleagues solved a HoxB1-Pbx1-DNA ternary complex at 2.35 A and concluded that heterodimer contacts are mediated by the HoxB1 hexapeptide binding a pocket in Pbx1 formed partly by a three-amino-acid insertion in the Pbx1 homeodomain, and that the Pbx1 DNA-binding domain is larger than the canonical homeodomain, containing an additional alpha helix that appears to contribute both to binding the HoxB1 hexapeptide and to stable DNA binding 2. Consistent with that, PDB 1B72 models the PBX1 chain over residues 233-319, twenty-four residues past UniProt's annotated homeodomain end of 295. The HOXA9-PBX1 complex 1PUF models 233-305.

Both are carried: data.domains contains the UniProt homeodomain 233-295 and a separate entry for the structurally observed extension to 319, flagged as such. Drawing only 233-295 would hide the helix the crystallographers specifically wrote a sentence about.

The break cuts PBC-A in half1 deeper

UniProt annotates the TCF3-PBX1 site at 88/89 1. That falls inside PBC-A (45-124), removing residues 45-88 and leaving 89-124. The array marks PBC-A lost on the grounds that half a subdomain is not a subdomain, and says so in the note rather than pretending a clean excision.

PBC-B (127-232) and the homeodomain (233-295, or 233-319 structurally) are untouched. Nourse and colleagues described the fusion as replacing the E2A DNA-binding domain with 'the putative DNA binding domain of a homeoprotein from chromosome 1' 4 — the domain map confirms that description precisely: what TCF3::PBX1 gains is an intact, HOX-cooperative, TALE homeodomain and the PBC-B surface that HOX hexapeptides dock into.

Which makes the fusion a HOX-cofactor with an activation domain welded on

Wild-type PBX1 is a positioning cofactor: it does not carry a strong activation domain of its own, and its job is to specify where HOX proteins bind. TCF3::PBX1 keeps the positioning machinery and adds TCF3's AD1 and AD2. The functional statement that follows straight from the coordinates is that the fusion converts a specificity module into an activation module aimed at PBX/HOX sites.

What is not annotated

InterPro returns no additional PBX1 entries beyond the PBC and homeobox matches, and UniProt annotates disorder at 1-40, 317-338 and 395-430. So residues 296-430 — a third of the protein, all of it retained in the fusion — have no domain assignment at all except for a single Site at 286 'required for binding to the NFIL3 promoter' 1. That region is drawn as bare ruler in the viewer, which is the correct depiction: we do not know what is there.

Facts and figures

uniprot	P40424
length AA	430
gene symbol	PBX1
isoform note	Coordinates are UniProt canonical P40424, 430 aa, which is the numbering used by the t(1;19) literature and by the structural papers. PBX1 has shorter splice variants but they are not the reference for fusion breakpoints.
assumed fusion	TCF3::PBX1, t(1;19). PBX1 contributes residues 89-430 per the UniProt-annotated breakpoint site at 88/89.
alphafold ID	AF-P40424-F1
domain coordinate caveat	PBC-A and PBC-B are UniProt sub-region assignments with no experimental structure behind them, so treat them as soft to tens of residues; the parent PBC homology domain 38-232 is the better-supported interval and is given here only in this caveat to avoid drawing three nested bars. The homeodomain start at 233 is solid (both crystal structures begin exactly there). The homeodomain END is genuinely disputed: UniProt 295, PDB 1PUF 305, PDB 1B72 319, with the structure paper stating explicitly that the domain is larger than a canonical homeodomain. Residues 296-430 other than the modelled helix have no annotation of any kind.
fusion boundary residue	89

domains

PDB IDS

[
  "1B72",
  "1PUF"
]

domains — the numbers behind the chart

[
  {
    "name": "PBC-A",
    "start": 45,
    "end": 124,
    "note": "TRUNCATED: the TCF3::PBX1 break at 88/89 removes 45-88 and leaves 89-124. Marked lost because roughly half the subdomain is gone. Part of the wider PBC homology domain 38-232.",
    "fate": "lost"
  },
  {
    "name": "PBC-B",
    "start": 127,
    "end": 232,
    "note": "retained intact; part of the wider PBC homology domain 38-232. Forms part of the pocket that HOX hexapeptides dock into.",
    "fate": "retained"
  },
  {
    "name": "HD (TALE homeodomain, DNA-binding)",
    "start": 233,
    "end": 295,
    "note": "UniProt's annotation. See the next entry: the crystallographic DNA-binding domain is larger.",
    "fate": "retained"
  },
  {
    "name": "HD extension (fourth helix, structural)",
    "start": 296,
    "end": 319,
    "note": "NOT ANNOTATED BY UNIPROT. Piper et al. showed the Pbx1 DNA-binding domain contains an additional alpha helix beyond the canonical homeodomain; PDB 1B72 models the chain to 319 (1PUF only to 305). Recorded separately because the sources genuinely disagree.",
    "fate": "retained"
  }
]

pbc full domain

{
  "start": 38,
  "end": 232,
  "note": "the parent PBC homology domain, of which PBC-A and PBC-B are the annotated halves"
}

How this could bite us back

PBX1 is a general HOX cofactor required in development; PBX1 haploinsufficiency in humans causes congenital anomalies of the kidney and urinary tract. Anything that blocks the HOX-PBX hexapeptide pocket hits normal developmental transcription, and the pocket is shared across the PBX family.
The fusion's DNA-binding surface is wild-type PBX1 surface, so there is no fusion-specific epitope on this side; selectivity must come from the TCF3 moiety or the junction.

Still open

Does the truncated PBC-A remnant (residues 89-124) fold onto PBC-B in the fusion, or does it dangle? This matters because PBC-A is part of the surface that regulates PBX1's own nuclear import and cofactor binding.
Where does the PBX1 DNA-binding domain really end? Two crystal structures of the same protein with different HOX partners model it to 305 and 319. A construct-truncation binding series would settle a boundary that currently varies by 24 residues.
Residues 296-430 are retained in the fusion and are completely unannotated. Is that third of the protein doing anything, and would deleting it in a fusion background change transformation?

Every source behind this entry (5)

E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
PBX1 is 430 residues with a PBC homology domain at 38-232 (PBC-A 45-124, PBC-B 127-232) and a TALE-type homeodomain at 233-295, and UniProt annotates the TCF3-PBX1 translocation site between residues 88 and 89.
UniProtKB entry P40424 (PBX1_HUMAN): reviewed sequence, length and positional feature annotations · UniProt Consortium · UniProtKB/Swiss-Prot 2026
What this evidence cannot show
The homeodomain end at 295 conflicts with the crystallographic evidence for a fourth helix extending to ~319 (see E2). UniProt annotates no feature at all in 296-430 other than a single functional site at 286.
Evidence that argues against this (1)
- PDB 1B72 models the PBX1 chain to residue 319 and the authors state that the Pbx1 DNA-binding domain is larger than the canonical homeodomain. Piper DE et al. 1999
E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The 2.35 A HoxB1-Pbx1-DNA ternary structure shows that the Pbx1 DNA-binding domain is larger than a canonical homeodomain, containing an additional alpha helix that contributes both to binding the HoxB1 hexapeptide and to stable DNA binding, with the hexapeptide pocket formed partly by a three-residue insertion in the Pbx1 homeodomain.
Structure of a HoxB1-Pbx1 heterodimer bound to DNA: role of the hexapeptide and a fourth homeodomain helix in complex formation · Piper DE, Batchelor AH, Chang CP, Cleary ML, Wolberger C · Cell 1999
PMID 10052460 DOI
What this evidence cannot show
A ternary complex of isolated domains on a short DNA duplex; PDB 1B72 models PBX1 residues 233-319 and 1PUF (HOXA9-PBX1) models 233-305, so even the structural extent of the extra helix differs by 14 residues between two structures. Both titles were retrieved and verified at RCSB.
Evidence that argues against this (1)
- That fourth helix does not exist until DNA is present. Multidimensional NMR of the enlarged Pbx1 DNA-binding domain shows the additional alpha helix is UNFOLDED when Pbx1 is free in solution and that its folding is triggered by Pbx1 binding DNA; no conformational change was seen on mixing HoxB1 with Pbx1 in the absence of DNA. The crystal's larger domain is therefore a DNA-induced state, and drawing helix 4 as a static feature of the PBX1 domain map, or reasoning about it in a fusion context without DNA, overstates the structure. Jabet C et al. 1999
E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
Domain databases return no PBX1 features beyond the PBC and homeobox matches, leaving residues 326-430 unannotated.
InterPro/Pfam/CATH-Gene3D/SUPERFAMILY domain matches for UniProt P40424 · InterPro Consortium; EMBL-EBI · InterPro 2026
What this evidence cannot show
Absence of annotation is weak evidence of absence of function; it means only that no profile model matches, which for an intrinsically disordered C-terminus is expected. CORRECTION 2026-09-19: this item previously said the unannotated stretch was 296-430. PANTHER PTHR11850 reaches 325, CATH-Gene3D G3DSA:1.10.10.60 reaches 315 and SUPERFAMILY SSF46689 reaches 301, so the genuinely unannotated region is 326-430. The figure was 29 residues too wide.
Evidence that argues against this (1)
- Three member databases do annotate past 296, so the unannotated stretch is 29 residues shorter than stated. The live InterPro call for P40424 returns PANTHER PTHR11850 (TALE homeobox family) at 36-325, CATH-Gene3D G3DSA:1.10.10.60 (Homeodomain-like) at 233-315 and SUPERFAMILY SSF46689 / InterPro IPR009057 at 233-301, all of which extend beyond 296. The genuinely unannotated region is 326-430, not 296-430 - and the structure-based methods, CATH and SUPERFAMILY, are the ones that run furthest, which matters because it is structural extent that a fusion breakpoint argument depends on. InterPro Consortium; EMBL-EBI 2026
E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The t(1;19) fusion replaces the E2A DNA-binding domain with the DNA-binding domain of a chromosome-1 homeoprotein (subsequently named PBX1), producing a chimeric transcription factor.
Chromosomal translocation t(1;19) results in synthesis of a homeobox fusion mRNA that codes for a potential chimeric transcription factor · Nourse J, Mellentin JD, Galili N, et al. · Cell 1990
PMID 1967982 DOI
What this evidence cannot show
1990 cDNA cloning in three cell lines; the PBX1 contribution was defined by sequence, not by a protein-level measurement.
Evidence that argues against this (1)
- The PBX1 homeodomain is not required for transformation. An E2a-Pbx1 construct lacking the homeodomain entirely (E2a-PbxdeltaHD) efficiently transformed NIH 3T3 fibroblasts and induced malignant lymphomas in transgenic mice; either the homeodomain OR the Pbx1 residues flanking it could mediate transformation when fused to E2a, while mutating the E2a transactivation motifs AD1 and AD2 significantly impaired it. The conclusion - transformation is absolutely dependent on E2a-derived motifs and the Pbx1 homeodomain is optional - inverts the emphasis of the chimeric-transcription-factor framing this item carries. Monica K et al. 1994
E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
An AlphaFold model AF-P40424-F1 covers all 430 residues.
AlphaFold Protein Structure Database entry for P40424 (model AF-P40424-F1) · Jumper J, Varadi M, et al.; EMBL-EBI/DeepMind · AlphaFold DB 2026
What this evidence cannot show
Experimental coverage exists only for 233-319; the PBC domain has no experimental structure in this entry's PDB cross-references, so the AlphaFold PBC-A/PBC-B geometry is unvalidated.
Evidence that argues against this (1)
- Nearly half the model is in the two lowest confidence bands. Per-residue pLDDT recomputed from the v6 file gives 136 of 430 residues (32%) below 50 and a further 57 (13%) between 50 and 70 - 45% of the chain at or below the caution threshold - with a mean of 69.2. Experimental coverage is thinner still: PDBe lists only two structures for P40424 in the entire PDB, 1B72 (233-319) and 1PUF (233-305), so 343 of 430 residues have never been seen by any method. AlphaFold DB / EMBL-EBI; PDBe SIFTS 2026

Connected to

refutes →TCF3::PBX1 (E2A-PBX1), t(1;19)(q23;p13.3)
Counter-evidence added cycle 41: the PBX1 homeodomain is DISPENSABLE for transformation. E2a-PbxdeltaHD, lacking the homeodomain entirely, efficiently transformed NIH 3T3 cells and induced malignant lymphomas in transgenic mice, while mutating the E2a AD1 and AD2 transactivation motifs significantly impaired it (Monica 1994, PMID 7969166). The DNA-binding-swap framing is right about the sequence and misleading about the mechanism.
← fuses withTCF3 (E2A) — domain architecture and residue coordinates
TCF3::PBX1, t(1;19)
part of →TCF3::PBX1 (E2A-PBX1), t(1;19)(q23;p13.3)
3-prime partner domain architecture; PBC and the TALE homeodomain are retained
supports →Mechanism class 2 — transcription-factor corruption (RUNX1::RUNX1T1, PML::RARA, ETV6::RUNX1, CBFB::MYH11) — a balance, not a pure dominant negative
the fusion swaps the regulated PBX1 N-terminus for a constitutive TCF3 activation domain
fuses with →TCF3 (E2A) — domain architecture and residue coordinates
t(1;19)(q23;p13.3)
supports →Mechanism class 5 — enhancer hijacking (inv(3) GATA2/EVI1 — and DUX4-r, which only half belongs here)
the retained TALE homeodomain supplies PBX/HOX site specificity to a foreign activation domain
related to →KMT2A / MLL1 — the gene that breaks
both act through HOX-cofactor transcriptional programmes in B-lineage leukaemia
← related toTCF3::PBX1 (E2A-PBX1), t(1;19)(q23;p13.3)
Counter-evidence: Monica 1994 (PMID 7969166) shows E2a-Pbx lacking the PBX1 homeodomain still transforms fibroblasts and gives lymphomas in mice, so 1B72's HOX-cooperative chemistry is not a fusion requirement
← related toHow HOXA9 binds DNA: a hexapeptide into a PBX pocket, a 1.9 A structure that is mouse HoxA9 plus human Pbx1 homeodomains, and a second route that makes the motif redundant
PDB 1PUF chain 4 is human PBX1 (P40424), the protein that entry covers, as a 73-residue homeodomain fragment
part of →TCF3::PBX1 (E2A-PBX1), t(1;19)(q23;p13.3)
3' partner domain coordinates