ASH2L SPRY and RBBP5 347-356: a phospho-reader contact gated by RbBP5 Ser350, and the only WRAD interface with a measured fold-change
also called ASH2L SPRY-RbBP5 · RbBP5 S350 phosphoswitch · 4X8N
Two of KMT2A's four helper proteins stick to each other only when a single serine on one of them carries a phosphate. Remove the phosphate and the pair falls apart, which shuts down the enzyme - in the normal protein, not in the fusion.
The same thing, for a scientist
PDB 4X8N (2.1 A) is the ASH2L SPRY domain bound to a phosphorylated RbBP5 peptide. The deposited RBBP5 entity ERESEFDIED aligns uniquely to Q15291 residues 347-356 against the fetched UniProt sequence, which places the phospho-acceptor Ser350 in the centre of the peptide and, per the authors, in the centre of the ASH2L SPRY concave surface. ITC shows the phosphorylated RbBP5 344-357 peptide binds ASH2L SPRY with 15-fold higher affinity than the unphosphorylated peptide, and Flag-RbBP5 S350A fails to co-immunoprecipitate ASH2L. S350 makes no significant direct contact itself - the phosphate does the work - which makes the ASH2L SPRY domain a phospho-reader. This is WRAD-internal chemistry and therefore wild-type-allele biology in a KMT2A-rearranged cell, but it is the sharpest example in the module of an interface that is conditional rather than constitutive.
Residues and the numbering
RBBP5 (Q15291): crystallised peptide 347-356 (ERESEFDIED), phospho-acceptor S350, within the D/E box; the ITC peptide is 344-357 12.
ASH2L (Q9UBL3): the SPRY domain concave surface; the deposited SPRY entity is a 181-residue fragment 2.
A numbering warning that applies to the whole ASH2L literature: the commonly used ASH2L construct and the papers built on it number from the 534-residue isoform, while UniProt Q9UBL3 is 628 residues. The offset is exactly 94. Any ASH2L residue index taken from a paper must be checked against which isoform it counts from - see the DPY30 entry, where the same offset relocates the entire SDI motif.
Why a conditional interface matters here
Most of the interfaces in this beat are constitutive: a motif either is or is not present in the fusion. Two are conditional on phosphorylation - this one, and the CK2-gated LEDGF IBM switch on the KMT2A N-terminus. The difference in consequence is total. The LEDGF switch sits on the fragment the fusion keeps, so it is a potential handle on the fusion itself. This one sits between two WRAD subunits the fusion never recruits, so the kinase that writes it is a modifier of the wild-type allele.
That asymmetry is worth stating explicitly because 'phospho-regulated interface in the MLL complex' reads as one category and is in fact two, with opposite therapeutic implications.
| partner a | gene-ash2l |
|---|---|
| partner b | gene-rbbp5 |
| motif b | RBBP5 D/E box, crystallised 347-356, ITC peptide 344-357, phospho-acceptor S350 |
| motif a | ASH2L SPRY domain concave surface (deposited as a 181-residue fragment) |
| kd status | no absolute Kd published in the accessible text; the measured quantity is a 15-fold ITC affinity increase on S350 phosphorylation |
| fate in fusion | contextually lost - both subunits exist in the cell but are recruited through the KMT2A WIN motif, which no der(11) fusion retains |
| numbering trap | ASH2L papers commonly number from the 534-residue isoform while UniProt Q9UBL3 is 628 residues; the offset is exactly 94 |
| druggability | a phospho-dependent peptide groove; the pharmacologically accessible node is the kinase, which is unidentified for S350 in the accessible text |
| resolution note | PDB metadata and entity sequences fetched from data.rcsb.org 2026-09-20; RBBP5 span aligned to UniProt Q15291 fetched the same day |
interface residues b
[ "RBBP5 S350 (phospho-acceptor; no significant direct contact itself)" ]
kd measurements
[
{
"ligand": "phospho-RbBP5 344-357 vs unphosphorylated, to ASH2L SPRY",
"fold_change": 15,
"direction": "tighter when phosphorylated",
"method": "ITC",
"source_pmid": "25593305"
}
]structures
[ "4X8N", "6PWV", "7UD5" ]
structure details
[
{
"pdb": "4X8N",
"title_fetched": "Crystal structure of Ash2L SPRY domain in complex with phosphorylated RbBP5",
"method": "X-ray",
"resolution_A": 2.1,
"contents": "ASH2L 181-mer SPRY + RBBP5 10-mer (Q15291 347-356)",
"fetched": "2026-09-20"
}
]- Wrong allele: this interface is assembled on wild-type MLL1, so disrupting it in a KMT2A-rearranged cell attacks the allele that is not driving the disease.
- Breadth: ASH2L-RBBP5 is shared by all SET1-family complexes, so a disruptor removes H3K4 methylation globally.
- Kinase-level intervention: targeting the writer of S350 would hit every other substrate of that kinase, and the kinase is not even identified.
- Assay artefact: the co-IP evidence uses ectopically expressed Flag-RbBP5, so stoichiometry in the assay is not physiological.
- Which kinase phosphorylates RbBP5 S350? The paper establishes the consequence and cites proteomic detection but the accessible text does not name the enzyme.
- Is RbBP5 S350 phosphorylation altered in KMT2A-rearranged cells, where the wild-type allele is the only MLL1 that can use WRAD at all?
- What is the absolute Kd of the phosphorylated and unphosphorylated RbBP5 peptide for ASH2L SPRY? Only the ratio is published.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Phosphorylation of RbBP5 Ser350 increases binding of the RbBP5 344-357 peptide to the ASH2L SPRY domain 15-fold by ITC, and RbBP5 S350A fails to co-immunoprecipitate ASH2L; S350 makes no significant direct contact with ASH2L in the structure.
A phosphorylation switch on RbBP5 regulates histone H3 Lys4 methylation. · Zhang P, Chaturvedi CP, Tremblay V, et al. · Genes Dev 2015What this evidence cannot show
The 15-fold is reported as a fold-change; no absolute Kd for either state appears in the accessible full text. The co-IP uses ectopically expressed Flag-RbBP5 in cells, and the functional readout is H3K4me3 at the beta-globin locus in murine erythroleukaemia cells, not a leukaemia model. Three prior proteomics studies are cited as evidence that S350 is phosphorylated in vivo.
Evidence that argues against this (1)
- An ASH2L Tyr359Val mutant in the same paper behaves like wild type, showing that not every residue in the SPRY concave surface is load-bearing and that mutational evidence at this interface needs the specific residue, not the region. Zhang P et al. 2015
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
PDB 4X8N is a 2.1 A crystal of the ASH2L SPRY domain with a phosphorylated RBBP5 peptide mapping to Q15291 347-356.
PDB 4X8N: Crystal structure of Ash2L SPRY domain in complex with phosphorylated RbBP5 · Zhang P, Chaturvedi CP, Tremblay V, et al. · RCSB Protein Data Bank 2015What this evidence cannot show
Title, method, resolution and both polymer entity sequences fetched from data.rcsb.org on 2026-09-20; the residue span is this cycle's alignment to the UniProt Q15291 sequence fetched the same day. The ASH2L entity is a 181-residue SPRY fragment and reports nothing about the rest of the protein.
Evidence that argues against this (1)
- In the assembled cryo-EM complexes ASH2L is present as a 534-residue chain with large intrinsically disordered regions that modulate activity on chromatin and are not in this crystal, so the isolated SPRY-peptide picture is partial. Lee YT et al. 2021
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
ASH2L and DPY30 intrinsically disordered regions modulate MLL/SET1 activity on chromatin and are largely unresolved in the assembled structures.
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
Mechanistic assignment from cryo-EM plus biochemistry on reconstituted complexes; the MLL1 component is again a C-terminal fragment.
Evidence that argues against this (1)
- Because the disordered regions are unresolved, their contribution is inferred from activity assays rather than seen, which is the same class of inference the graph flags elsewhere as an interface asserted without residue-level mapping. Rahman S et al. 2022
- binds →ASH2L — the WRAD subunit that has DPY30's essentiality profile and none of its co-dependency signal
SPRY domain reads phospho-S350 on RBBP5
RBBP5 347-356, phospho-acceptor S350
the other end of RBBP5; together they chain WDR5 to ASH2L
a phospho-gated interface, but on WRAD rather than on the fragment the fusion keeps
residue-level detail for the wild-type side
the other ASH2L contact in WRAD; both carry the same 94-residue isoform trap