Lost from the KMT2A half: the four-residue HCF-binding motif (DHNY, aa 1799-1802)
also called HBM · HCF-1 binding motif MLL · D/EHXY motif · MLL 1799-1802
Normal KMT2A carries a four-letter tag that hooks it to a control protein called HCF-1. That tag sits just past the place where the gene always breaks, so every leukemic fusion protein loses it.
The same thing, for a scientist
HCF-1 (HCFC1) and HCF-2 (HCFC2) join the 1-MDa wild-type MLL complex by binding a single conserved tetrapeptide in MLL that fits the (D/E)HXY HCF-binding-motif consensus. Yokoyama et al. map it to MLL amino acids 1799-1802 and show that a deletion mutant (Delta-1799/1803) loses HCF-1 and HCF-2 while retaining ASH2L and WDR5 binding [E1]. Independent inspection of UniProt Q03164 finds DHNY at exactly 1799-1802 and no other [DE]Hx Y match anywhere in the 3,969-residue sequence, so the motif's position is not in doubt [E2]. Because the KMT2A breakpoint cluster region sits around exons 8-13 (roughly residues 1,100-1,400), residue 1799 is C-terminal to every reported translocation breakpoint: the HBM is absent from every KMT2A fusion protein. This is the cleanest-defined subtraction on the KMT2A side of the fusion - four residues, one consensus, one binding partner - and it is also the one with the least functional follow-up.
Why this is the cleanest lost module in the whole protein1 deeper
Most 'lost domain' claims about KMT2A fusions are claims about folded units whose boundaries vary by tens of residues between sources, so 'lost' really means 'the fold is disrupted somewhere in here'. The HBM is different. It is a four-residue linear motif with a published consensus, a published position, and a published deletion mutant that separates its loss from the loss of anything else in the same region 1.
Two independent lines put it at 1799-1802:
- Yokoyama et al. state the residue numbers directly and note the sequence 'fits a consensus (D/EHXY) for the previously defined HBM'; their Delta-1799/1803 mutant loses HCF-1 and HCF-2 but still binds ASH2L and WDR5 1.
- Scanning the UniProt Q03164 sequence for the [DE]Hx Y pattern returns exactly one hit in 3,969 residues: DHNY at 1799-1802 2. A four-residue pattern with roughly a 1-in-4,000 expectation per site matching once, at the published position, is about as good as motif confirmation gets without new experiments.
The breakpoint cluster region spans roughly residues 1,100-1,400. Residue 1799 is beyond it by ~400 residues, and beyond the PHD1-PHD3 cassette as well. No reported KMT2A fusion retains it.
The same paper is the source of the project's founding observation
This is the study the graph elsewhere quotes for 'oncogenic mutant forms of MLL retain an ability to interact with menin but not other identified complex components'. Read in full rather than from the abstract, the paper says two things that matter here and are usually collapsed into one:
- Menin binding maps within the first 1,406 amino acids of MLL-N - i.e. inside the retained fragment, which is why menin survives the translocation 1.
- Menin was the only identified component of the MLL/HCF complex that associates with mutant MLL proteins 1.
So the menin-inhibitor strategy and the HBM subtraction are two faces of the same measurement: the breakpoint falls between the menin site and everything else the complex needs. The retained-versus-lost boundary is not an accident of which proteins were looked for; it is a position on a ruler.
What is NOT known: whether losing it matters1 deeper
This entry is a subtraction with a confirmed position and no demonstrated functional consequence in leukemia. Specifically:
- No one has re-inserted the HBM into a KMT2A fusion and asked whether the fusion is impaired. (Compare PHD3, where exactly that experiment exists -
subtraction-kmt2a-phd3-bromo-cassette.) - HCFC1 is a near-universal cell-essential gene in DepMap, depleted in effectively all lines, which means an HCFC1-directed strategy would have no therapeutic window and also means HCFC1 dependency in KMT2A-r cells tells you nothing specific about the fusion.
- HCF-1 has many partners beyond MLL (E2F factors, THAP11, ZNF143, OGT), so the phenotype of HCF-1 loss is not the phenotype of MLL-HCF-1 uncoupling.
The honest label is: lost, precisely located, consequence unmeasured. That is worth recording as such rather than upgrading to 'a brake the fusion had to shed'.
The experiment this motif makes unusually cheap
A four-residue insertion is a far better add-back reagent than a 60-residue PHD finger. The standing objection to the PHD3 re-insertion result is that inserting a folded domain into a fusion protein could impair it sterically rather than by restoring a switch. Inserting DHNY - or the minimal Yokoyama fragment - into KMT2A::AFF1 carries almost none of that risk, and its negative control (a single H-to-A substitution that kills HCF binding while preserving length) is trivial. See exp-hbm-reinsertion-kmt2a-aff1.
| module | HCF-binding motif (HBM) |
|---|---|
| uniprot | Q03164 |
| start | 1799 |
| end | 1802 |
| sequence | DHNY |
| consensus | (D/E)HXY |
| fate in fusion | lost |
| position relative to bcr | C-terminal to the breakpoint cluster region (exons 8-13, approx. residues 1100-1400) by roughly 400 residues |
| separation of function mutant | Delta-1799/1803 (loses HCF-1/HCF-2, retains ASH2L and WDR5) |
| demonstrated functional consequence of loss | none reported |
| loss class | lost, precisely located, consequence unmeasured |
position confirmed by
[ "primary source full text (PMID 15199122, via PMC480881)", "UniProt Q03164 canonical sequence regex scan (unique [DE]HxY match)" ]
binding partners lost
[ "HCFC1 (HCF-1, P51610)", "HCFC2 (HCF-2, Q9Y5Z7)" ]
- HCFC1 is essentially pan-essential in DepMap, so any therapeutic idea routed through HCF-1 itself has no window; only the MLL-HCF interface would be targetable, and that interface is already absent from the fusion, which means disrupting it can only harm the wild-type allele the leukemic cell still carries.
- If the wild-type KMT2A allele is required in KMT2A-rearranged cells, then a drug that breaks the HBM-HCF-1 contact would act on the allele the cell needs, not on the oncogene - the opposite of selectivity.
- Does re-inserting the four-residue HBM into KMT2A::AFF1 restore HCF-1 recruitment, and does that change anything about the fusion's program or its transforming capacity? The reagent is trivial to build and the experiment has never been done.
- Wild-type MLL is cleaved by taspase1 into MLL-N (residues 1-2666) and MLL-C; the HBM at 1799 therefore stays with MLL-N in wild-type cells but is absent from the fusion. Does HCF-1 distinguish the wild-type MLL-N fragment from the fusion at shared target loci?
- HCF-1 is essential in nearly every cell line. Is there any KMT2A-r-selective consequence of MLL-HCF uncoupling that could be read out without an HCFC1 knockdown, e.g. by a separation-of-function HCF-1 allele that cannot bind the HBM?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
HCF-1 and HCF-2 are components of the 1-MDa wild-type MLL complex and bind a conserved motif at MLL amino acids 1799-1802 fitting the (D/E)HXY HBM consensus; a Delta-1799/1803 mutant loses HCF-1 and HCF-2 binding while retaining ASH2L and WDR5; menin interaction maps within the first 1,406 amino acids of MLL-N; and menin is the only identified component of the MLL/HCF complex that associates with mutant (fusion) MLL proteins.
Leukemia proto-oncoprotein MLL forms a SET1-like histone methyltransferase complex with menin to regulate Hox gene expression · Yokoyama A, Wang Z, Wysocka J, Sanyal M, Aufiero DJ, Kitabayashi I, Herr W, Cleary ML · Molecular and Cellular Biology 24(13):5639-5649 2004What this evidence cannot show
FULL TEXT read via PMC (PMC480881) for the residue numbers, the Delta-1799/1803 mutant, the 1,406-residue menin mapping and the 'only component' statement; the abstract alone gives none of the residue numbers. Biochemistry in transfected 293 cells and in HB cells expressing endogenous MLL-ENL, from 2004: complex composition was defined by what the purification and the available antibodies could see, so 'the only component retained' means 'the only one of the components identified in this study'. Later work added components (e.g. the PAF complex, LEDGF/PSIP1) that this survey did not test. No leukemia phenotype is attached to HBM loss anywhere in the paper.
Evidence that argues against this (1)
- The 'menin is the only retained component' framing should not be read as 'the fusion complex is otherwise empty'. Menin physically links MLL to LEDGF/PSIP1, LEDGF is required for MLL-dependent transcription and for leukemic transformation, and LEDGF was not among the components surveyed in 2004 - so the retained interactome of the fusion is larger than the founding paper could see. ABSTRACT ONLY. Yokoyama A et al. 2008
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The UniProt Q03164 (KMT2A_HUMAN, primary gene name KMT2A, 3,969 aa) sequence contains the tetrapeptide DHNY at residues 1799-1802, and a regular-expression scan of the full sequence for the HBM consensus [DE]Hx Y returns this as the only match in the protein.
UniProtKB entry Q03164 (KMT2A_HUMAN) - canonical sequence and feature table, retrieved via the UniProt REST API · UniProt Consortium · UniProtKB 2026What this evidence cannot show
Accession resolved by fetching Q03164 directly and reading back the PRIMARY gene name (KMT2A) and length (3,969 aa), not by a gene_exact search, which matches synonyms. The regex scan was run by this Queen on the downloaded canonical FASTA; it is arithmetic on a public sequence, not an experiment, and it confirms only the position and uniqueness of the motif, not that HCF binds it. UniProt does NOT annotate an HCF-binding motif feature at this position, so the annotation here comes from E1 and the sequence match, not from UniProt's own curation.
Evidence that argues against this (1)
- A regular-expression search of one canonical sequence establishes that DHNY is the unique [DE]Hx Y match in that isoform. It does not establish functional uniqueness: HCF-1 binds through an (D/E)HxY consensus that tolerates context, motif matches in other isoforms or in the reciprocal product were not searched, and sequence uniqueness in a 3,969-residue protein is a weak prior for biological uniqueness. Molecular and Cellular Biology 2004
the HBM route is established for HCF-1 and only inferred for HCF-2
- part of →KMT2A / MLL1 — the gene that breaks
HCF-binding motif DHNY at aa 1799-1802, lost in every fusion
- related to →HCFC1 (HCF-1) — a founding member of the wild-type KMT2A complex that the fusion cannot bind
the binding partner the fusion can no longer recruit
same measurement, opposite side of the breakpoint: menin binds within aa 1-1406 and is retained; HCF binds at 1799 and is not
module absent from this and every other KMT2A fusion protein
loss follows from breakpoint position: residue 1799 lies ~400 aa past the cluster region
row: HBM 1799-1802
- ← testsExperiment: put four residues back - the minimal add-back control the PHD3 result has never had
first functional test of whether HBM loss matters to a fusion