FLI1 — domain architecture and residue coordinates
also called FLI1 · Friend leukemia integration 1 · EWSR2 · SIC-1 · Q01543
FLI1 is a DNA-gripping transcription factor with two working parts: a self-stacking module near the front and a DNA-reading module near the back. In Ewing sarcoma the self-stacking module is discarded and only the DNA-reading module is kept, welded onto a sticky region from EWSR1.
The same thing, for a scientist
FLI1 (UniProt Q01543, 452 aa) is an ETS-family transcription factor with a PNT/SAM domain at 112-198 and an ETS DNA-binding domain at 281-361. In the type 1 EWSR1::FLI1 fusion, EWSR1 exon 7 joins FLI1 exon 6; mapping the Ensembl FLI1 canonical coding exons places the end of exon 5 at codon 218, so the FLI1 contribution begins at approximately **residue 219** and runs to 452. The PNT domain is therefore entirely lost and the ETS domain entirely retained: the mirror image of what happens to ETV6 in ETV6::RUNX1, where the SAM domain is kept and the ETS domain discarded. UniProt's ETS boundary (281-361) is considerably narrower than every crystal construct, which span 259-399, and that disagreement is recorded.
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.
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.
Two ETS proteins, opposite surgery
Put the two ETS-family comparators side by side and the contrast is exact:
| ETV6 in ETV6::RUNX1 | FLI1 in EWSR1::FLI1 type 1 | |
|---|---|---|
| PNT/SAM domain | retained (40-124) | lost (112-198) |
| ETS DNA-binding domain | lost (339-420) | retained (281-361) |
| Role in the fusion | supplies oligomerisation | supplies DNA targeting |
Same protein family, same two domains, and the breakpoint selects the opposite half in each case. Whatever the partner cannot supply, the ETS protein provides: RUNX1 brings its own DNA-binding Runt domain, so ETV6 contributes polymerisation; EWSR1 brings no DNA-binding activity at all, so FLI1 contributes the ETS domain. That is a clean, verifiable illustration of mechanism-different-breaks-different-biology and it is drawn straight from the coordinates.
The junction, computed
Type 1 is defined as EWS exon 7 joined in frame to FLI1 exon 6 3. We mapped the Ensembl FLI1 canonical coding exons and found coding exon 5 ends at codon 218 (+1 nt) and exon 6 at codon 240 4. FLI1 exon 6 therefore begins at the second nucleotide of codon 219, and the FLI1 contribution to the fusion is conventionally written 219-452.
UniProt annotates no translocation site for FLI1 at all, so unlike BCR, TCF3, PBX1, PML, ETV6 and CBFB there is no independent curated coordinate to check this against. Confidence is accordingly 'strong' rather than 'established', and the entry is marked approximate. Type 2 (EWS exon 7 to FLI1 exon 5) would begin near residue 197 and includes more FLI1; the ETS domain is retained in both.
The ETS boundary is annotated too narrowly
Every FLI1 ETS-domain structure is wider than UniProt's 281-361 2:
| Structure | FLI1 chain extent |
|---|---|
| 1FLI ('DNA-BINDING DOMAIN OF FLI-1', NMR) | 276-373 |
| 5E8G, 5E8I | 276-399 |
| 5JVT, 6VG2, 6VG8, 6VGD | 276-375 |
| 9CP6 | 259-371 |
| 9MWY, 9MX9, 9MXA | 259-399 |
The structural consensus is that the DNA-binding unit runs from about 259-276 to about 371-399, i.e. it is flanked on both sides by helices that UniProt's profile-based 281-361 excludes. The array records UniProt's 281-361 as the core and adds the flanking extent in the note, because the flanks are where the additional DNA contacts and the partner interfaces sit. This is the same pattern seen with PBX1's fourth helix: profile methods find the recognition helix bundle, crystallographers find a larger functional unit.
| uniprot | Q01543 |
|---|---|
| length AA | 452 |
| gene symbol | FLI1 |
| isoform note | Coordinates are UniProt canonical Q01543, 452 aa, matching the Ensembl canonical transcript FLI1-203 used for the exon arithmetic. Alternative isoform lengths verified directly: 386, 419 and 259 aa. The Ewing sarcoma literature numbers against the 452 aa form, so there is no material numbering trap here, unlike ABL1, RUNX1 and NUP98. |
| assumed fusion | EWSR1::FLI1 type 1 (EWSR1 exon 7 to FLI1 exon 6). FLI1 contributes approximately residues 219-452, computed from Ensembl exon coordinates. Type 2 (to FLI1 exon 5) would begin near residue 197; the ETS domain is retained in both. |
| alphafold ID | AF-Q01543-F1 |
| domain coordinate caveat | The PNT domain at 112-198 is well supported (the NMR construct 1X66 covers 114-198). The ETS domain is the problem entry: UniProt and InterPro give 281-361 from ETS profile models, while nine independent structures span 259-399, so the profile annotation is 20 residues short at the N-terminal end and up to 38 short at the C-terminal end. Both are recorded. The fusion boundary at ~219 is computed from Ensembl exon coordinates and has NO independent curated cross-check because UniProt annotates no FLI1 breakpoint, so it is the least corroborated junction coordinate in this comparator set and is marked approximate. |
| fusion boundary residue | 219 |
| fusion boundary confidence | approximate: computed from exon arithmetic only, with no independent UniProt breakpoint annotation to check it against. The junction codon is chimeric. |
PDB IDS
[ "1FLI", "1X66", "5E8G", "6VGD" ]
domains
[
{
"name": "PNT (SAM/pointed domain)",
"start": 112,
"end": 198,
"note": "entirely lost in the type 1 fusion (junction ~219). PDB 1X66 covers 114-198, closely matching the annotation. The mirror image of ETV6, whose PNT domain is what its fusion keeps.",
"fate": "lost"
},
{
"name": "ETS (DNA-binding domain)",
"start": 281,
"end": 361,
"note": "UniProt/InterPro profile boundary. Every crystal and NMR construct is wider, spanning 259-399 across entries; the additional flanking helices carry DNA contacts. Treat 281-361 as the core and 259-399 as the functional unit.",
"fate": "retained"
}
]- The retained ETS domain is wild-type sequence with a DNA-binding surface shared across the large ETS family, so any DNA-competitive agent risks hitting ETS1, ERG, FLI1 itself and dozens of others.
- FLI1 is required for megakaryocyte development and haematopoiesis; germline FLI1 variants cause a bleeding disorder, so degrading FLI1 alongside the fusion has a known human phenotype waiting at the other end.
- Because the fusion's DNA-binding module is entirely wild-type FLI1 and its activation module is entirely wild-type EWSR1, the only genuinely fusion-specific feature is the junction itself, which is a short disordered stretch and a poor target for anything except a T-cell receptor or antibody approach.
- Where does the FLI1 ETS domain functionally end? Profile databases say 361, structures go to 399, and the difference covers a helix that contacts DNA. A truncation series measuring DNA affinity at 361, 375 and 399 would resolve a 38-residue ambiguity in the single most important domain of this fusion.
- The type 1 breakpoint at ~219 is the only fusion junction in this comparator set with no independent curated corroboration. Mapping published type 1 cDNA junctions onto the canonical coding sequence would fix that cheaply.
- ETV6 and FLI1 are both ETS proteins with the same two-domain layout, and their fusions keep opposite halves. Is that determined entirely by what the partner lacks, or does something about the intron structure of each gene make one breakpoint more accessible than the other?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
FLI1 is 452 residues with a PNT/SAM domain at 112-198 and an ETS DNA-binding domain at 281-361, with predicted disorder at 209-271 and 433-452, and no annotated translocation site.
UniProtKB entry Q01543 (FLI1_HUMAN): reviewed sequence, length and positional feature annotations · UniProt Consortium · UniProtKB/Swiss-Prot 2026What this evidence cannot show
The absence of any curated breakpoint site for FLI1 means the fusion boundary in this node has no independent UniProt cross-check, unlike most other proteins in this set. UniProt lists four isoforms with verified lengths 452, 386, 419 and 259.
Evidence that argues against this (1)
- Every other domain database draws these two boxes somewhere else. For PNT the live InterPro call returns SMART SM00251 114-198, Pfam PF02198 115-197, CDD cd08541 114-204, SUPERFAMILY SSF47769 99-197, CATH-Gene3D G3DSA:1.10.150.50 103-200 and InterPro IPR013761 99-200 - a start spread of 16 residues (99 to 115) and an end spread of 7 (197 to 204) against UniProt's 112-198. For ETS the spread is far worse: Pfam PF00178 280-360, SMART SM00413 280-365, InterPro IPR000418 280-365, SUPERFAMILY SSF46785 278-413, CATH-Gene3D G3DSA:1.10.10.10 247-373 and InterPro IPR036390 278-413. The ETS domain therefore runs anywhere from 247 to 413 depending on method, and the stored 281-361 is one of eight coexisting answers rather than a consensus. InterPro Consortium; EMBL-EBI 2026
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Crystal and NMR structures of the FLI1 DNA-binding domain model chains over 276-373 (1FLI), 276-399 (5E8G), 259-371 (9CP6) and 259-399 (9MWY), all wider than the UniProt ETS annotation of 281-361.
PDB 1FLI 'DNA-BINDING DOMAIN OF FLI-1', 5E8G 'Crystal structure of the DNA binding domain of human transcription factor FLI1' and 6VGD 'Crystal structure of the DNA binding domain (DBD) of human FLI1 and the complex of the DBD of human Runx2 with core binding factor beta (Cbfb), in complex with 16mer DNA' (all titles retrieved and verified at RCSB) · RCSB Protein Data Bank (entry metadata retrieved and title verified) · RCSB PDB 2026What this evidence cannot show
Construct extents reflect what crystallised, not necessarily a domain boundary; but the consistency of the ~259-399 window across independent groups over thirty years is stronger than any single construct.
Evidence that argues against this (1)
- UniProt and InterPro both give the narrower 281-361, derived from ETS profile models rather than from structures. UniProt Consortium 2026
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Type 1 EWS-FLI1, the commonest form, joins EWS exon 7 in frame to FLI1 exon 6.
EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewing's sarcoma · de Alava E, Kawai A, Healey JH, et al. · Journal of Clinical Oncology 1998What this evidence cannot show
Exon nomenclature for FLI1 counts the first coding exon as exon 1 in the Ensembl model used here; a source that counts a non-coding first exon would be off by one. ADDED 2026-09-19: the structural claim (EWS exon 7 to FLI1 exon 6) is what this item is carried at 'established' for. The prognostic claim in the same source did NOT replicate in Euro-E.W.I.N.G. 99 (565 informative patients, PMID 20308673) and must not be cited from here.
Evidence that argues against this (1)
- The prognostic half of the cited paper failed to replicate prospectively. Euro-E.W.I.N.G. 99 typed 703 primary tumours and analysed 565 informative patients (type 1 n=296, type 2 n=133, other EWS-FLI1 n=91, EWS-ERG n=45) with 4.5-year median follow-up and found no significant prognostic value of fusion type; the only non-null signal was a hazard ratio of 1.38 (95% CI 0.96-2.0, P=0.10) for the non-type-1/non-type-2 group, and the authors state that the prospective evaluation did not confirm a prognostic benefit for type 1 EWS-FLI1 fusions. The structural fact - EWS exon 7 to FLI1 exon 6 - stands; the prognostic framing the cited paper carries does not. Le Deley MC et al. 2010
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Recomputing the Ensembl FLI1 canonical coding exon structure (ENST00000527786, 452 aa) places the end of coding exon 5 at codon 218 and exon 6 at codon 240, so a fusion joining to FLI1 exon 6 begins at approximately residue 219.
Ensembl transcript ENST00000527786 (FLI1): exon structure and coding-sequence coordinates, used to convert exon breakpoints into residue numbers · Ensembl; EMBL-EBI · Ensembl REST API 2026What this evidence cannot show
Computed by us. Exon 5 ends mid-codon (+1 nt), so the junction codon is chimeric. Because UniProt annotates no FLI1 breakpoint, this figure has no independent curated corroboration and should be treated as approximate.
Evidence that argues against this (1)
- Exon 6 is not a stable address in Ensembl's own FLI1 model. The live gene record lists 12 protein-coding FLI1 transcripts whose exon counts run 4, 5, 7, 7, 8, 8, 8, 8, 9, 9, 10, 10 and whose protein lengths run 475, 459, 452, 432, 430, 430, 419, 410 and shorter; the longest, FLI1-211 / ENST00000897157, has 10 exons and 475 aa against the chosen canonical's 9 exons and 452 aa. A breakpoint reported as FLI1 exon 6 in a paper built on a 10-exon model converts to a different residue than 219, and nothing in the primary literature pins which model was used. Ensembl; EMBL-EBI 2026
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
EWS-FLI1 recruits BAF complexes to tumour-specific enhancers as a neomorphic property that wild-type FLI1 does not have.
Cancer-Specific Retargeting of BAF Complexes by a Prion-like Domain · Boulay G, Sandoval GJ, Riggi N, et al. · Cell 2017What this evidence cannot show
The comparison with wild-type FLI1 is the key control; it establishes that the FLI1 moiety alone does not retarget BAF, so the FLI1 contribution is DNA targeting rather than cofactor recruitment.
Evidence that argues against this (1)
- The GGAA-microsatellite selectivity the tumour-specific-enhancer account rests on is an intrinsic property of the bare FLI1 DNA-binding domain, not something the fusion creates. Crystal structures of two, three and four FLI1 DBD molecules on DNA carrying two, three and four contiguous GGAA sites show the isolated DBD binding contiguous GGAA repeats cooperatively and forming a nucleoprotein filament. Cooperative microsatellite occupancy therefore does not require the EWS prion-like domain, and a BAF-recruitment result cannot distinguish 'EWS retargets BAF' from 'FLI1 already occupies microsatellites cooperatively and EWS supplies the activating surface'. Hou C et al. 2025
- E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
An AlphaFold model AF-Q01543-F1 covers all 452 residues, and experimental structures exist for both the PNT domain (1X66, 114-198) and the ETS domain.
AlphaFold Protein Structure Database entry for Q01543 (model AF-Q01543-F1) · Jumper J, Varadi M, et al.; EMBL-EBI/DeepMind · AlphaFold DB 2026What this evidence cannot show
The region between the two domains (roughly 199-258) is predicted-disordered with no experimental coverage, and it is precisely the region the type 1 breakpoint falls in.
Evidence that argues against this (1)
- Covers is a coordinate count, not information. Per-residue pLDDT recomputed from the v6 model gives 219 of 452 residues (48%) below 50 - the band AlphaFold itself labels likely disordered - with only 135 residues (30%) at or above 90 and a mean of 63.4. The PNT structure is also weaker than 'experimental structure' implies: 1X66 is a solution NMR entry with no resolution, and a second NMR entry 2YTU covers a wider 100-220 construct. The ETS-domain crystal structures (9CP6 at 1.66 A, 9MWY, 9MX8, 9MX9, 9MXA, 5E8G, 5JVT) span 259-399 in SIFTS, and 9CP6 chain A leaves auth residues 255-271 and 371 unobserved, so its real coverage is 272-370. AlphaFold DB / EMBL-EBI; PDBe SIFTS 2026
Counter-evidence added cycle 41: cooperative GGAA-microsatellite occupancy is intrinsic to the bare FLI1 DNA-binding domain, which forms a nucleoprotein filament on contiguous GGAA repeats without any EWS prion-like domain (Hou 2025, PMID 40131773). A BAF-retargeting result therefore cannot distinguish 'EWS retargets' from 'FLI1 already binds and EWS supplies the activating surface'.
EWSR1::FLI1, t(11;22)
3-prime partner domain architecture; the ETS DNA-binding domain is retained and the PNT domain lost
- analogous to →ETV6 — domain architecture and residue coordinates
two ETS-family proteins entering fusions from opposite ends: FLI1 contributes its ETS domain, ETV6 its PNT domain
- ← analogous toETV6 — domain architecture and residue coordinates
same ETS-family two-domain layout, opposite halves retained by their fusions
t(11;22)(q24;q12), type 1
- supports →Do different breaks give different biology? A verdict on partner identity vs breakpoint position
FLI1 keeps its ETS domain and loses PNT, the exact mirror of ETV6 in ETV6::RUNX1
the retained ETS domain supplies the DNA targeting that directs the EWSR1 EAD to GGAA microsatellites
- ← related toEWSR1::FLI1 (EWS-FLI1), t(11;22)(q24;q12)
Counter-evidence: Sorensen 1994 (PMID 8162068) identified t(21;22) fusing identical EWS sequence to ERG, so the ETS partner is interchangeable and t(11;22) covers only 85% of cases