NUP98::TOP1, t(11;20)(p15;q11~q12) — the most precisely mapped NUP98 breakpoint, and the reciprocal that was wrong twice
also called NUP98-TOP1 · t(11;20)(p15;q11) · t(11;20)(p15;q12) · NUP98/TOP1
This fusion has the best-defined join of any in the family: the same two exons are stitched together in every reported patient. It was first thought to happen only after chemotherapy and to make only one of its two possible products — both of those beliefs turned out to be wrong.
The same thing, for a scientist
t(11;20)(p15;q11), later written **q11~q12**, joins the NUP98 **FXFG repeats** to the body of **DNA topoisomerase I** (P11387, 765 aa), retaining the Topo IB catalytic domain (**432–765**) including the catalytic **Tyr723** [E1][E5]. The junction is the most precisely specified in the family: **NUP98 exon 13 to TOP1 exon 8, in all ten cases reported in the literature** [E4], with genomic breaks in **NUP98 intron 13 and TOP1 intron 7** [E2]. That exon-13 junction sits exactly where the family's breakpoints cluster — 44.9% of all NUP98 fusions break in exon 13 [E6]. The founding report framed the lesion as **therapy-related MDS** and described the breakpoint as separating *"the N-terminal FXFG repeats from the RNA-binding domain located in the C-terminus"* of NUP98, with **no reciprocal transcript** [E1]. Both framings were subsequently overturned: a **de novo** AML with no chemical exposure was reported together with the **first detection of the reciprocal TOP1/NUP98 transcript** [E2]; t(11;20) was then argued to identify a subset of **non-therapy-related** AML associated with FAB M2, younger age, hyperleukocytosis and poor prognosis [E3]; and the reciprocal was found coexpressed in **4 de novo AML cases but 0 of 3 therapy-related MDS cases**, suggesting the reciprocal tracks progression to AML [E4].
The exon junction, and why it is the useful fact
NUP98 exon 13 → TOP1 exon 8, in all ten reported cases 4; genomic breaks in NUP98 intron 13 and TOP1 intron 7 2.
That is a diagnostic-grade fact. A single junction across every reported patient means a single RT-PCR assay detects the lesion, and it means the retained NUP98 fragment is the same in every patient — so any structure-function claim about the NUP98 side generalises across this partner without qualification, which is not true for partners with variable breakpoints.
It also anchors the family-level breakpoint statistic: 39.7% of NUP98 fusions break in exon 12 and 44.9% in exon 13, about 85% in those two exons combined 6. NUP98::TOP1 sits in the larger of the two bins, and an exon-13 break retains FG repeats 1, GLEBS and FG repeats 2 — the standard retained module across the family.
Two claims the founding paper got wrong, and what that teaches
- 'Therapy-related.' The founding cases were therapy-related MDS 1. Two later reports found de novo disease 23, and one argued the translocation actually marks a non-therapy-related AML subset with distinctive features — FAB M2, younger age, hyperleukocytosis, poor prognosis 3. The founding context was ascertainment, not biology. The same correction had to be made for NUP98::HOXD13.
- 'No reciprocal.' The founding paper reported the reciprocal TOP1/NUP98 absent 1; it was later detected 2, and then found in 4 of 4 de novo AML cases and 0 of 3 therapy-related MDS cases 4, which if it holds makes the reciprocal a marker of progression rather than a passenger.
The transferable lesson for this graph is about negative claims in founding papers. 'No reciprocal transcript detected' in 1999 meant 'not detected by the assay used on the samples available'. The same phrasing appears in the NUP98::DDX10 founding paper and in the NUP98::PRRX1 report, and in at least this one case it did not survive. A negative from a founding cytogenetic paper should be carried at low confidence by default.
The retained catalytic domain, and an untested implication
The fusion retains the Topo IB-type catalytic domain (P11387 residues 432–765) including the catalytic Tyr723 5. So unlike NUP98::DDX10, where the annotated catalytic function turned out to be a passenger (fusion-nup98-ddx10), NUP98::TOP1 plausibly carries an active enzyme.
Nobody has tested whether the catalytic activity is required for transformation — no Tyr723Phe fusion construct appears in the literature found here. That is the same gap the DDX10 case exposed, and it carries a specific therapeutic implication that has not been pursued: topoisomerase I inhibitors already exist and are used in paediatric oncology. If the fusion's TOP1 activity is required, a camptothecin-class agent is an obvious and immediately available hypothesis. If it is not required — the DDX10 outcome — then the partner is supplying an interaction surface and the inhibitor would be beside the point. One construct separates these.
A caution in the other direction: topoisomerase II inhibitors are the classic cause of therapy-related NUP98 rearrangement, so the relationship between this fusion and topoisomerase-directed drugs is already tangled.
| gene5 | NUP98 |
|---|---|
| gene3 | TOP1 |
| cytoband | t(11;20)(p15;q11~q12) |
| partner class | topoisomerase |
| uniprot5 | P52948 |
| uniprot3 | P11387 |
| length AA 5 | 1817 |
| length AA 3 | 765 |
| reciprocal product | TOP1/NUP98; reported absent by the founding paper, later detected, and coexpressed in 4 de novo AML versus 0 of 3 therapy-related MDS |
| catalytic activity required | UNTESTED; no Tyr723Phe fusion construct reported |
| therapeutic hypothesis not pursued | if the retained TOP1 catalytic activity is required, topoisomerase I inhibitors (camptothecin class) are an immediately available hypothesis |
| menin dependency status | NUP98-TOP1 was one of the five fusions shown to bind MLL1 and NSL by targeted pull-down, but it has not been tested against a menin inhibitor |
| prognosis | poor in the small reported series; no cohort data; age range includes an 81-year-old |
cytoband variants in literature
[ "t(11;20)(p15;q11)", "t(11;20)(p15;q12)" ]
diseases
[ "therapy-related MDS", "de novo AML (FAB M2)" ]
breakpoints
[
{
"id": "bp-nup98-top1-exon13-exon8",
"gene5_exon_last": 13,
"gene3_exon_first": 8,
"frequency_pct": 100,
"population": "all ten reported NUP98::TOP1 cases",
"domains_retained_5": [
"FG repeats 1 (1-156)",
"GLEBS (157-213)",
"FG repeats 2 (214-480)"
],
"domains_lost_5": [
"peptidase S59 (738-880)",
"Nup96 chain (881-1817)"
],
"domains_retained_3": [
"Topo IB-type catalytic domain (432-765)",
"catalytic Tyr723"
],
"domains_lost_3": [
"TOP1 N-terminal domain upstream of exon 8"
],
"phenotype_notes": "Genomic breaks in NUP98 intron 13 and TOP1 intron 7; the single most reproducible junction in the NUP98 family."
}
]- Topoisomerase II inhibitors are the classic cause of therapy-related NUP98 rearrangement, so proposing a topoisomerase-directed therapy for a topoisomerase fusion needs an explicit argument about why the class that helped cause these leukaemias should now treat one.
- A perfectly reproducible exon junction makes a clean diagnostic assay but also means every patient has the identical retained protein, so there is no natural variation to learn from and structure-function must come from constructs.
- This partner binds MLL1 in the assay that underpins the family's menin rationale, yet has never been exposed to a menin inhibitor; enrolling such a patient on that rationale would be extrapolation from a pull-down.
- Is the retained TOP1 catalytic activity required for transformation? A Tyr723Phe fusion construct would answer it, and a positive answer would put an approved drug class in play for this partner.
- Is the reciprocal TOP1/NUP98 functional, or is its association with de novo AML an artefact of improving assay sensitivity over two decades?
- Does NUP98::TOP1 respond to menin inhibition? It binds MLL1 in the pull-down but has never been tested against a menin inhibitor in any model.
- Why is this junction so reproducible when other NUP98 partners vary? Intron 13 of NUP98 and intron 7 of TOP1 may share a recombination-prone feature worth looking for.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
t(11;20)(p15;q11) associated with therapy-related myelodysplastic syndrome results in a NUP98-TOP1 fusion; the breakpoint separates the N-terminal FXFG repeats of NUP98 from its C-terminal RNA-binding region, producing a chimeric mRNA encoding the NUP98 FXFG repeats fused to the body of DNA topoisomerase I.
The t(11;20)(p15;q11) chromosomal translocation associated with therapy-related myelodysplastic syndrome results in an NUP98-TOP1 fusion · Ahuja HG, et al. · Blood 94(9):3258-3261 1999What this evidence cannot show
A 1999 case report; PubMed holds no DOI for this record so none is given. Two of its framings - therapy-related only, and no reciprocal transcript - were subsequently overturned. The 'RNA-binding domain' it describes in the NUP98 C-terminus is not annotated as an RRM in UniProt P52948, so no residue boundaries for it are given in this graph.
Evidence that argues against this (1)
- The therapy-related framing did not hold: a de novo AML with no chemical exposure history was subsequently reported, and a later series argued t(11;20) identifies a subset of non-therapy-related AML. Potenza L et al. 2004
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A de novo AML with t(11;20)(p15;q11) and no chemical exposure history expressed both NUP98/TOP1 and the reciprocal TOP1/NUP98 transcript - the first detection of a reciprocal the founding paper had reported absent - with genomic breakpoints in NUP98 intron 13 and TOP1 intron 7.
Both NUP98/TOP1 and TOP1/NUP98 transcripts are detected in a de novo AML with t(11;20)(p15;q11) · Iwase S, et al. · Genes, Chromosomes & Cancer 38(1):102-105 2003What this evidence cannot show
A single case. Detection of a reciprocal transcript says nothing about whether it is translated or functional, and RT-PCR sensitivity differences between 1999 and 2003 are a sufficient explanation for the discrepancy with the founding report.
Evidence that argues against this (1)
- A reciprocal transcript present in one patient and absent in the founding series may simply reflect assay sensitivity rather than a biological difference between patients, and no functional role for TOP1/NUP98 has ever been demonstrated. Ahuja HG et al. 1999
- E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
t(11;20)(p15;q11) may identify a subset of non-therapy-related acute myelocytic leukaemia associated with FAB M2, younger age, hyperleukocytosis and poor prognosis.
A t(11;20)(p15;q11) may identify a subset of nontherapy-related acute myelocytic leukemia · Potenza L, et al. · Cancer Genetics and Cytogenetics 2004What this evidence cannot show
The title's hedge - 'may identify' - is appropriate and is retained here. The claim rests on a small number of reported cases assembled across reports rather than a cohort, and 'younger age' is not quantified in a way that could be compared with the family-level age distributions.
Evidence that argues against this (1)
- A later series reports an 81-year-old case, so the age range for this fusion is wide and 'younger age' as a defining feature is not secure. Yamamoto K et al. 2016
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In all ten cases described in the literature, the NUP98 exon 13 / TOP1 exon 8 fusion transcript was expressed; the reciprocal TOP1/NUP98 transcript was coexpressed in 4 cases of de novo AML but 0 of 3 therapy-related MDS, suggesting the reciprocal is associated with progression to AML.
Coexpression of NUP98/TOP1 and TOP1/NUP98 in de novo Acute Myeloid Leukemia with t(11;20)(p15;q12) and t(2;5)(q33;q31) · Yamamoto K, et al. · Cytogenetic and Genome Research 150(3-4):287-292 2016What this evidence cannot show
A literature review within a case report; 4 versus 0 of 3 is seven patients in total and the comparison is retrospective across reports using different assays over nearly two decades. The index case also carried a second translocation, t(2;5)(q33;q31), which confounds it.
Evidence that argues against this (1)
- Detection of a reciprocal transcript across reports separated by two decades is confounded by improving assay sensitivity: later cases are more likely to be de novo AND more likely to have the reciprocal detected, which would produce this association with no biology behind it. Iwase S et al. 2003
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
TOP1 is P11387, 765 aa, with no UniProt gene-name synonyms, and the fusion retains the Topo IB-type catalytic domain at residues 432-765 including the catalytic Tyr723.
UniProtKB entry P11387, DNA topoisomerase 1 (TOP1) · UniProt Consortium · UniProtKB 2026What this evidence cannot show
Primary gene name and length read back from the fetched record. That the fusion retains this region follows from the exon-8 junction and the domain annotation together; the two were not read from a single source, so the retention claim is an inference from two verified facts.
Evidence that argues against this (1)
- Retaining a catalytic domain does not establish that the catalytic activity is used. The NUP98::DDX10 case is the cautionary precedent in this very family: the annotated enzymatic function was a passenger and a 24-residue interaction surface did the work. No catalytically dead NUP98-TOP1 construct has been reported. Shima Y et al. 2025
- E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Across paediatric NUP98-rearranged AML, about 85% of NUP98 breakpoints fall in exon 12 (39.7%) or exon 13 (44.9%), with exon 14 breakpoints almost exclusively NUP98::KDM5A.
Comprehensive molecular and clinical characterization of NUP98 fusions in pediatric acute myeloid leukemia · Bertrums EJM, et al. · Haematologica 108(8):2044-2058 2023What this evidence cannot show
Derived from a paediatric cohort in which NSD1 and KDM5A dominate; NUP98::TOP1 is one of the rare 'NUP98-X' partners and contributes at most a case or two to this distribution, so the family-level statistic is not independent evidence about this fusion's breakpoint.
Evidence that argues against this (1)
- A breakpoint distribution dominated by two partners may not describe rare partners at all. The NUP98::TOP1 exon-13 junction is established from its own ten-case series, not from this distribution, and the agreement between the two is corroboration rather than derivation. Yamamoto K et al. 2016
- analogous to →NUP98::DDX10, inv(11)(p15q22) — the 'helicase' fusion whose helicase activity turns out not to be the point
THE TWO NON-HOMEODOMAIN NUP98 FUSIONS, and the pair that argues the partner's ENZYME is not what transforms. Both fuse the NUP98 FXFG/GLFG repeats to a catalytic protein - a topoisomerase and a putative RNA helicase - and in both the catalytic function turns out not to be the point: for DDX10 the transforming element is a 24-residue NOL10-binding stretch, not the helicase. Both arose in therapy-related MDS/AML. Together they make the FG repeats the constant and the enzyme the accident.
- part of →NUP98 — domain architecture and residue coordinates (with the 1817/920 numbering trap resolved)
NUP98 exon 13 to TOP1 exon 8 in all ten reported cases
one of the five KMT2A-free fusions in the MLL1/NSL pull-down
its exon-13 junction sits in the family's dominant breakpoint bin
3' partner coordinates; Topo IB catalytic domain 432-765 retained intact
contributes the 5' half: the FG/GLFG region and the off-pore chromatin-facing module