DPY30 D/D domain and the ASH2L SDI helix: four hydrophobic pockets, one load-bearing leucine - and a 94-residue numbering offset that relocates the whole motif
also called ASH2L SDI · DPY-30 dimerization/docking domain · Sdc1-DPY-30 interacting region · 4RIQ
DPY30 works as a pair and grips a single short helix from ASH2L in a cradle made of four small pockets. One leucine does most of the work, and swapping it out breaks the grip and lowers the histone mark the complex writes.
The same thing, for a scientist
PDB 4RIQ (2.23 A) is the DPY-30 dimerization/docking (D/D) domain, residues 45-99, bound to a 16-residue ASH2L SDI helix. Two DPY-30 protomers build a cradle of four pockets: P-I contacts ASH2L V509, P-II binds T512 and L513, P-III binds V516 and L517, and P-IV binds V520 (all in the paper's numbering). Alanine substitution of the polar residues H511 and D515 does not impair binding, and aspartate substitution at V509 or V520 only modestly does, but L513D cripples it - and L513D reduces H3K4me3 and beta-globin expression in differentiating erythroleukaemia cells. The residue indices in that paper count from the 534-residue ASH2L isoform; against UniProt Q9UBL3 (628 aa) every number is 94 higher, so the SDI helix crystallised in 4RIQ is Q9UBL3 603-618 and the load-bearing leucine is Q9UBL3 L607.
The numbering offset, because it will bite someone
The 4RIQ ASH2L entity sequence, with its expression tag and TEV site removed, is VEHTLADVLYHVETEV. That string occurs exactly once in the fetched UniProt Q9UBL3 sequence, at 603-618. The paper calls the same fragment 509-524. The difference is a constant 94 and is fully explained by ASH2L isoform choice: the widely used ASH2L construct is 534 residues, UniProt canonical is 628, and 628 - 534 = 94. The cryo-EM deposits 6PWV and 7UD5 also carry 534-residue ASH2L chains, so the short numbering is the field's default.
Every ASH2L residue index in this graph should therefore be qualified by isoform. Recorded here in both systems:
| paper (534-aa isoform) | UniProt Q9UBL3 (628 aa) | role |
|---|---|---|
| V509 | V603 | P-I pocket |
| H511 | H605 | polar, dispensable |
| T512 | T606 | P-II |
| L513 | L607 | P-II, load-bearing |
| D515 | D609 | polar, dispensable |
| V516 | V610 | P-III |
| L517 | L611 | P-III |
| V520 | V614 | P-IV |
DPY30's side
DPY-30 (Q9C005, 99 residues) contributes residues 45-99 as the D/D domain - a fold with striking homology to the dimerisation/docking domain of protein kinase A regulatory subunits. The pockets are built from both protomers: P-II from Arg54(A), Leu57(A), Val62(A), Leu65(B) and Leu69(B); P-III from Val62(B), Leu57(B), Leu69(A), Leu65(A) and Leu66(A) 1. The interface is therefore obligate on DPY-30 dimerisation, which is the structural reason a monomeric DPY30 could not substitute and a reason this contact is hard to drug with a small molecule: half of each pocket belongs to the other copy.
Relation to the fusion
None of this reaches KMT2A::AFF1. DPY30 and ASH2L reach MLL1 through WDR5 and RBBP5, which reach MLL1 through the WIN motif at 3762-3767 - absent from every der(11) fusion. The graph's hyp-dpy30-positions-ash2l-catalyses and hyp-wrad-reaches-kmt2a-fusions-only-as-a-second-shell-passenger are the right frames for what this interface can and cannot say about the disease, and this entry supplies their missing structural coordinates.
| partner a | gene-dpy30 |
|---|---|
| partner b | gene-ash2l |
| motif a | DPY-30 dimerization/docking (D/D) domain, Q9C005 residues 45-99, acting as a dimer |
| motif b | ASH2L Sdc1-DPY-30 interacting region (SDI); crystallised 509-524 in 534-aa isoform numbering = Q9UBL3 603-618 |
| kd status | no dissociation constant reported; the published evidence is structural plus overlay/pull-down mutagenesis and a cellular H3K4me3 readout |
| fate in fusion | contextually lost - reached only through the WIN motif the fusion lacks |
| obligate dimer | each ASH2L-binding pocket is built from residues of both DPY-30 protomers, so the interface presupposes DPY-30 dimerisation |
| druggability | poor as drawn: a short amphipathic helix in a composite groove spanning two protein copies, with no reported chemical matter and a receptor that also accepts NURF helices |
| resolution note | PDB metadata and entity sequences fetched from data.rcsb.org 2026-09-20; ASH2L mapping recomputed against UniProt Q9UBL3 fetched the same day after the published numbering failed to match |
interface residues a
[ "DPY30 R54", "DPY30 L57", "DPY30 V62", "DPY30 L65", "DPY30 L66", "DPY30 L69" ]
interface residues b
[ "ASH2L L513 (= Q9UBL3 L607), load-bearing", "ASH2L T512 (= T606)", "ASH2L V516 (= V610)", "ASH2L L517 (= L611)", "ASH2L V509 (= V603)", "ASH2L V520 (= V614)" ]
kd measurements
[]
structures
[ "4RIQ", "6PWV", "7UD5" ]
structure details
[
{
"pdb": "4RIQ",
"title_fetched": "Crystal structure of DPY-30 dimerization/docking domain in complex with Ash2L Sdc1-DPY-30 Interacting region (SDI)",
"method": "X-ray",
"resolution_A": 2.231,
"contents": "DPY30 56-mer (Q9C005 45-99) + ASH2L 27-mer including a TEV site; stripped peptide maps to Q9UBL3 603-618",
"fetched": "2026-09-20"
}
]numbering offset
{
"ash2l_paper_isoform_length": 534,
"uniprot_Q9UBL3_length": 628,
"offset": 94,
"worked_example": "paper L513 = UniProt L607"
}- Composite site: half of each binding pocket belongs to the second DPY-30 protomer, so a ligand must either accept an obligate dimer or break the dimer, and breaking it removes DPY-30 from NURF as well.
- Promiscuous receptor: the D/D pocket accommodates different amphipathic helices including NURF's, so selectivity is not available from the pocket shape.
- Wrong allele: the module is recruited through a KMT2A motif the fusion does not have.
- Numbering error risk: a 94-residue isoform offset sits between the published mutants and UniProt, and a mutagenesis plan built on the wrong system will target the wrong residues.
- Is there any measured affinity for DPY-30 D/D to the ASH2L SDI helix? None was found in this cycle, which leaves the one WRAD subunit with a favourable DepMap profile structurally defined and thermodynamically uncharacterised.
- Does the same DPY-30 pocket bind an NURF helix with comparable affinity? If so, a DPY-30 D/D ligand cannot be COMPASS-selective.
- Does L513D (Q9UBL3 L607D) affect KMT2A-rearranged cells, or only the erythroid MLL2-dependent programme in which it was tested?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The DPY-30 D/D domain (residues 45-99) binds ASH2L residues 509-524 in a four-pocket cradle built from both DPY-30 protomers; L513D cripples binding while H511A, D515A, V509D and V520D do not, and L513D reduces H3K4me3 and beta-globin expression in differentiating MEL cells.
Molecular basis for DPY-30 association to COMPASS-like and NURF complexes. · Tremblay V, Zhang P, Chaturvedi CP, et al. · Structure 2014What this evidence cannot show
Residue numbers are in the 534-residue ASH2L isoform, 94 lower than UniProt Q9UBL3. The functional readout is the beta-globin locus in murine erythroleukaemia cells under ASH2L knockdown and rescue - an erythroid differentiation model, not a leukaemia model, and the paper attributes the activity to MLL2 rather than MLL1. Full text reached by efetch db=pmc after fullTextXML returned HTTP 500.
Evidence that argues against this (1)
- The same authors show DPY-30 D/D accommodates different amphipathic helices and is used by NURF as well as COMPASS-like complexes, so an ASH2L-directed reading of this pocket over-specifies it; the pocket is a general helix receptor. Tremblay V et al. 2014
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
PDB 4RIQ is a 2.23 A crystal of DPY-30 with an ASH2L SDI peptide whose sequence maps to UniProt Q9UBL3 residues 603-618.
PDB 4RIQ: Crystal structure of DPY-30 dimerization/docking domain in complex with Ash2L Sdc1-DPY-30 Interacting region (SDI) · Tremblay V, Zhang P, Chaturvedi CP, et al. · RCSB Protein Data Bank 2014What this evidence cannot show
Title, method, resolution and both entity sequences fetched from data.rcsb.org on 2026-09-20; the Q9UBL3 mapping is this cycle's alignment to the UniProt sequence fetched the same day, after the paper's 509-524 numbering failed to match canonical UniProt. The deposited ASH2L entity includes a TEV recognition sequence that must be stripped before alignment.
Evidence that argues against this (1)
- The cryo-EM deposits 6PWV and 7UD5 carry 534-residue ASH2L chains, i.e. the short numbering, so a reader comparing a crystal to a cryo-EM map without checking isoform will get a consistent-looking but wrong correspondence. Rahman S et al. 2022
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DPY30 and ASH2L intrinsically disordered regions outside this interface modulate the complex's activity on chromatin.
Mechanism for DPY30 and ASH2L intrinsically disordered regions to modulate the MLL/SET1 activity on chromatin. · Lee YT, Ayoub A, Park SH, et al. · Nat Commun 2021What this evidence cannot show
Reconstituted complexes and cryo-EM; the disordered regions are inferred from activity rather than resolved.
Evidence that argues against this (1)
- If activity depends substantially on unresolved disordered regions, a crystal of the ordered D/D-SDI contact describes assembly but not regulation. Tremblay V et al. 2014
- binds →DPY30 — the smallest COMPASS subunit, and the only one that behaves like a selective dependency
D/D domain 45-99 as an obligate dimer
- binds →ASH2L — the WRAD subunit that has DPY30's essentiality profile and none of its co-dependency signal
SDI helix, Q9UBL3 603-618, load-bearing L607
supplies the structural coordinates that hypothesis lacked
the other ASH2L contact in WRAD; both carry the same 94-residue isoform trap
residue-level mapping for the wild-type side
the module's internal geometry, none of which contacts the fusion directly
both use the DPY30 dimerisation surface; whether AKAP95 and ASH2L compete for it has never been asked