Raise a KMT2A::AFF1 junction-specific monoclonal by the published TEL 6F2 protocol - rat immunisation with a junction peptide, ELISA and western screen, then a ChIP test against the wild-type-parent controls that the ETV6::RUNX1 campaign never ran
also called KMT2A::AFF1 junction antibody campaign
An antibody that recognises only the join between the two fused proteins would see the cancer protein and nothing else, and the recipe for making one against a childhood leukaemia fusion was published in 2013. This is that recipe applied to the fusion this project is about, with the specificity controls the original campaign left out.
The same thing, for a scientist
The feasibility objection to a KMT2A-fusion junction antibody has been that breakpoints are distributed. `absence-no-junction-specific-antibody-exists-for-any-kmt2a-fusion` already refutes that at the protein level: mRNA splicing collapses the genomic distribution onto **six** in-frame KMT2A::AFF1 protein junctions, all sharing the AFF1 acceptor octamer `QTYSNEVH`, because AFF1 exon 4 is the acceptor in 1,261 of 1,262 cases. **The remaining objection was that nobody has shown a junction peptide can raise a usable reagent against a chimeric transcription factor. That objection is now answered by a published precedent in a paediatric leukaemia fusion** [E1]: clone TEL 6F2, a 12-mer immunogen `PIGRIAECILGM` spanning the ETV6::RUNX1 fusion site, validated by ELISA and western blot, used for flow cytometry and for ChIP-on-chip. **The KMT2A::AFF1 immunogens are 16-mers, longer than the one that worked.** This experiment is that protocol, run against the commonest junction first, with the three controls the original lacked.
Design
Immunogen. Start with the commonest junction epitope of the six, HRIRVDFK|QTYSNEVH, which this graph's own computation already identifies as charged, exposed and in a disordered region. Couple to KLH and to ovalbumin, as in the precedent.
Host and fusion. Lou/C rats, subcutaneous and intraperitoneal, peptide-KLH with CpG oligonucleotide and incomplete Freund's; boost at 6 weeks without adjuvant; fuse spleen cells with P3x63-Ag8.653 three days later. This is the published protocol verbatim 1.
Screen. ELISA against the immunising 16-mer, then against the two 8-mer halves separately. A clone that binds either half alone is discarded - this is the control the ETV6::RUNX1 campaign did not report and it is the whole point of a junction reagent.
Validation, in order of increasing difficulty.
- Western blot on SEM and RS4;11 (KMT2A::AFF1) versus REH and a KMT2A-wild-type B-ALL line.
- Western blot on lysates overexpressing full-length KMT2A and full-length AFF1 separately and together - the mixture control that distinguishes junction recognition from bivalent binding of two adjacent epitopes.
- Immunoprecipitation, then ChIP against an N-terminal KMT2A antibody arm in the same cells.
The N-terminal arm is not optional. A junction antibody cannot see protein-kmt2a-star-truncated-product-of-the-internal-promoter or the wild-type allele, so a ChIP with the junction reagent alone measures the fusion's sites with no comparator; run both and report the difference.
What each outcome buys
- A clone that works on chromatin. The attribution problem this project has recorded across many entries - every published KMT2A-fusion occupancy dataset uses an antibody raised against retained sequence and therefore cannot separate fusion from wild-type - becomes solvable for one fusion. That is the largest single methodological gain available to this field.
- A clone that works on blots but not on chromatin. Still useful: junction-level detection of the fusion protein for residual-disease work, and a negative that is worth reporting because the ETV6::RUNX1 precedent claims the chromatin step and nobody has repeated it.
- No usable clone. A failed immunisation against a charged, exposed, disordered 16-mer is itself informative and should be published, because it is currently assumed rather than known that this is hard.
| design | Peptide immunisation of Lou/C rats with a KMT2A::AFF1 junction 16-mer conjugated to KLH and ovalbumin, CpG plus incomplete Freund's adjuvant, boost at 6 weeks, hybridoma fusion with P3x63-Ag8.653, ELISA screen against the full 16-mer and against each 8-mer half separately, then western, IP and ChIP validation against wild-type-parent controls. |
|---|---|
| readout | A monoclonal that detects KMT2A::AFF1 by western in SEM and RS4;11 and not in REH or a KMT2A-wild-type B-ALL line, does not bind either 8-mer half alone, and recovers known fusion target loci by ChIP above an N-terminal-KMT2A comparator. |
| cost tier | med |
| timeline | roughly 4-6 months to a screened clone on the published schedule (immunise, boost at 6 weeks, fuse at 6 weeks + 3 days, screen and subclone), plus 2-3 months of validation |
| target epitope | HRIRVDFK|QTYSNEVH - the commonest of six computed in-frame KMT2A::AFF1 protein junctions, sharing the AFF1 acceptor octamer QTYSNEVH |
| epitope set size | 6 |
| precedent immunogen length AA | 12 |
| this immunogen length AA | 16 |
| precedent pmid | 24121163 |
| precedent rrid | RRID:AB_3713372 |
| what it would close | absence-no-junction-specific-antibody-exists-for-any-kmt2a-fusion, and it would make fusion-versus-wild-type attribution decidable in occupancy experiments for one fusion |
| absence class | never-measured |
| absence class basis | No such campaign has been run for any KMT2A fusion, so nothing exists to be found in any archive. The Antibody Registry returns 0 for MLL-AF4, MLL-AF9, MLL-ENL, KMT2A-MLLT3 and KMT2A-AFF1 with the SS18-SSX clone passing as a same-template positive control (2026-09-22). |
controls
[ "each 8-mer half of the junction alone, by ELISA - the specificity control the precedent did not report", "co-expressed full-length KMT2A and full-length AFF1 in the same lysate, to exclude bivalent binding of two adjacent epitopes", "REH and a KMT2A-wild-type B-ALL line as negative cell controls", "an N-terminal KMT2A antibody arm run in parallel through every assay", "clone TEL 6F2 (RRID:AB_3713372) as a positive control for the ChIP step, IF it can be obtained" ]
falsifiers
[ "No hybridoma from a campaign of the published design binds the 16-mer in ELISA, or every binder also binds one 8-mer half alone. That would show the KMT2A::AFF1 junction is not immunogenic as a linear epitope and would convert the absence from 'nobody tried' to 'it does not work', which is a publishable negative.", "A clone passes ELISA and western but recovers no known KMT2A::AFF1 target locus by ChIP above input in SEM, while clone TEL 6F2 run in the same hands does recover ETV6::RUNX1 targets. That localises the failure to this epitope rather than to the approach." ]
- A junction antibody adopted as a diagnostic would score negative in any patient whose breakpoint produces a seventh, rarer junction. The six-junction figure is a coverage estimate over 1,262 published cases, not a guarantee, and a false negative in this disease is a missed high-risk assignment.
- In ChIP, an epitope occluded by a partner protein at the loci where the fusion is most engaged would yield a clean, reproducible map of the sites where the fusion is LEAST active, and nothing about the data would look wrong. The N-terminal comparator arm exists to catch this and must be reported even when it agrees.
- A reagent that sees only the fusion removes the wild-type allele and the truncated internal-promoter product from every experiment it is used in, which narrows an attribution problem by discarding the comparator rather than by resolving it.
- Is clone TEL 6F2 still available from Helmholtz Munich? Without it the ChIP arm has no positive control and a negative result cannot be attributed.
- Which of the six junction 16-mers has the best predicted linear B-cell epitope score? That is a desk calculation that would choose the first immunogen on evidence rather than on frequency alone.
- Would a junction antibody cross-react with the reciprocal AFF1::KMT2A product? The reciprocal has a different junction sequence by construction, so it should not - but that prediction has never been tested for any fusion and the reciprocal is present in a substantial fraction of t(4;11) cases.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
A junction-specific monoclonal against a paediatric leukaemia fusion oncoprotein was raised from a 12-mer peptide spanning the fusion site, in Lou/C rats with KLH conjugation and CpG adjuvant, screened by ELISA and western blot, and used for flow cytometry and ChIP-on-chip.
The impact of TEL-AML1 (ETV6-RUNX1) expression in precursor B cells and implications for leukaemia using three different genome-wide screening methods. · Blood Cancer Journal 2013What this evidence cannot show
Read in full 2026-09-22. The precedent is one campaign, one fusion, one laboratory, and the paper reports the successful clone without stating how many hybridomas were screened or how many failed - so it gives a protocol and not a success rate. ETV6::RUNX1 has a single dominant junction; KMT2A::AFF1 has six, so a full reagent set is six campaigns, not one.
Evidence that argues against this (1)
- The published specificity test is induced-versus-uninduced expression of the fusion in one engineered line. That cannot exclude a clone recognising a partner-side epitope whose conformation changes in the chimera, which would look identical in every assay reported and would fail immediately against a wild-type-parent overexpression control. This experiment therefore adds that control rather than copying the validation. Blood Cancer Journal 2013
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The reagent is catalogued and therefore citable and in principle obtainable: RRID:AB_3713372, clone TEL_6F2, non-commercial, applications FACS, Western Blot and ChIP.
Antibody Registry record 3713372 (RRID:AB_3713372), TEL-AML1 (ETV6-RUNX1) Antibody, clone TEL_6F2 · SciCrunch Antibody Registry 2026What this evidence cannot show
Fetched 2026-09-22 via the JSON API; the human-facing page is a client-rendered shell. The record carries zero citations and no availability field, so obtaining the existing clone as a positive control for a new campaign requires contacting the depositing laboratory and may not succeed.
Evidence that argues against this (1)
- If the ETV6::RUNX1 clone cannot be obtained, this experiment loses its best positive control - a junction antibody known to work in ChIP, run alongside the new one in the same hands. Without it, a ChIP failure cannot be attributed between the reagent and the protocol. SciCrunch Antibody Registry 2026
the experiment that would close this absence, using a protocol published for another paediatric leukaemia fusion
the protocol, the immunogen-length precedent and the ChIP positive control all come from this reagent
a reagent that detects this fusion and neither parent, which would make fusion-versus-wild-type attribution decidable in occupancy experiments