Oncofusion

start / graph / hyp-dpy30-positions-ash2l-catalyses

Hypothesisspeculative confidence 1 of 5: Our own inference or analogy. Not shown to be true.Q5c-interactome2

DPY30 aims the H3K4 machinery while ASH2L powers it, which is why only DPY30 tracks menin

also called WRAD division of labour · positioning versus catalysis hypothesis

Four proteins form the shared core of the enzyme that marks active genes, and only one of them, DPY30, is needed in the same cells that need menin. We think that is because DPY30's job is aiming the enzyme at the right genes, which is exactly what a KMT2A fusion held in place by menin requires.

The same thing, for a scientist

Within WRAD, DPY30 contributes only modestly to allosteric stimulation of KMT2 catalysis but its ASH2L interface governs genome-wide placement of H3K4me2/3, whereas ASH2L and RBBP5 engage the MLL1 SET-domain surface that licenses catalysis. We hypothesise that this division of labour explains why DPY30 is the only WRAD subunit whose leukemia dependency correlates with MEN1: a cell whose survival depends on aiming H3K4 methylation at a menin-tethered, non-default locus set needs the positioning subunit specifically.

Go deeper
The observation this is built to explain

Under an identical leukemia-restricted protocol across 54 DepMap leukemia lines, DPY30 correlates with MEN1 at r = 0.446 and survives Holm correction, while WDR5 (0.004), RBBP5 (-0.081) and ASH2L (-0.153) all have confidence intervals crossing zero 1.

The standard explanation in this graph — pan-essential genes have flat profiles and flat profiles cannot correlate — covers WDR5 (98.3% of lines depleted) and RBBP5 (97.6%) and fails for ASH2L, which is depleted in only 21.4% of lines, a distribution close to DPY30's 31.5%. ASH2L has the variance to correlate and does not. Something other than statistical power separates them.

Also relevant: DPY30's correlation with MEN1 is 0.11 across the full 1,208-line panel and 0.45 restricted to leukemia. The effect is contingent on cell context.

The mechanism proposed

Haddad and colleagues found that DPY30 contributes only modestly to positive heterotypic allosteric regulation of KMT2 methyltransferase activity in vitro, while an extensive DPY30–ASH2L interface network is critical for correct genome-wide placement of H3K4me2 and H3K4me3. Residual H3K4me2 after DPY30 loss sits preferentially in regions of highly transcribed genes 2.

Meanwhile RBBP5 makes the direct contact with H2B-conjugated ubiquitin that orients the complex on the nucleosome, and the RBBP5/ASH2L heterodimer engages a non-active-site surface of the MLL1 SET domain required for catalysis 34.

So: ASH2L and RBBP5 are catalysis hardware; DPY30 is the aiming mechanism. A cell that only needs bulk H3K4 methylation at genes that are already running hot needs the hardware. A cell that needs H3K4 methylation installed at a specific, otherwise-silent locus set — a KMT2A-fusion cell whose target selection is enforced by the menin–LEDGF tether — needs the aiming mechanism as well, and it needs it in proportion to how much it needs the tether.

That predicts exactly the correlation observed, and predicts it only in leukemia.

Why this matters beyond DPY30

If true, it generalises. It says that in any oncogenic chromatin complex whose pathology is mis-targeting rather than over-activity, the targeting subunits will show selective dependency and the catalytic subunits will show pan-essentiality. That is a rule for reading dependency data across fusion families, and it would sharpen the graph's standing conclusion that the hijacked machinery has no therapeutic window: the machinery's engine has no window, but its steering might.

If false — if the DPY30 correlation is noise in 54 lines, or a lineage artefact — then the most promising new entry in this beat evaporates, and so does one of the ten survivors of the cycle-9 correction.

Facts and figures
statementWithin the WRAD module, DPY30 functions primarily as a targeting/positioning subunit while ASH2L and RBBP5 function primarily as catalysis hardware; consequently DPY30, and not its complex-mates, is selectively required in cells whose survival depends on aiming H3K4 methylation at a menin-tethered locus set.
origindata-pattern
prior probability0.3
cost to testlow-to-medium: one structure-guided mutant, one CUT&RUN time course, one reanalysis of existing DepMap data
predictions
[
  "An ASH2L separation-of-function mutant that retains RBBP5 scaffolding but loses the DPY30 interface will phenocopy DPY30 loss in a KMT2A::AFF1 line, and will not phenocopy full ASH2L loss.",
  "DPY30 depletion in a KMT2A::AFF1 line will reduce H3K4me2/me3 preferentially at fusion-bound loci rather than globally, while RBBP5 depletion will reduce it globally.",
  "The DPY30-MEN1 co-dependency will survive restriction to B-ALL lines and separately to AML lines; if it is a lineage artefact it will evaporate in both.",
  "Menin inhibition and DPY30 depletion will be epistatic rather than additive in a KMT2A::AFF1 line, because both act on the same targeting step.",
  "Across other fusion families whose pathology is mis-targeting rather than over-activity, targeting subunits will show selective dependency while catalytic subunits show pan-essentiality."
]
experiments
[
  "exp-ash2l-dpy30-interface-separation-of-function"
]
analyses
[
  "ana-wrad-subunit-discrimination"
]
falsifiers
[
  "The ASH2L DPY30-interface mutant behaves like full ASH2L loss rather than like DPY30 loss, which would mean the two subunits are not functionally separable in cells.",
  "DPY30 depletion reduces H3K4me2/3 uniformly across the genome in a KMT2A::AFF1 line rather than preferentially at fusion-bound loci.",
  "The DPY30-MEN1 correlation disappears when the leukemia panel is split by lineage, indicating a lineage confounder rather than shared mechanism.",
  "DPY30 depletion and menin inhibition prove additive rather than epistatic, indicating they act on separate steps.",
  "A larger leukemia panel returns a DPY30-MEN1 correlation whose confidence interval crosses zero, retracting the observation the hypothesis was built to explain."
]
How this could bite us back
  • If the hypothesis is right, the target it implies is still DPY30, a 99-residue protein whose only function is an extended helical interface and whose conditional mouse knockout causes severe pancytopenia with an HSC differentiation block.
  • A targeting subunit shared by six complexes has no selectivity handle even if the mechanistic story is correct.
  • The hypothesis is fitted to an observation made in 54 cell lines with a confidence interval reaching down to 0.14. It could be explaining noise.
Still open
  1. Does the positioning-versus-catalysis distinction survive the fact that the MLL1 SET domain, which ASH2L and RBBP5 bind, is absent from the fusion? If WRAD reaches fusion biology only through the wild-type allele, the hypothesis needs restating in those terms.
  2. Is there a third possibility, that DPY30's correlation with MEN1 reflects a function outside SET1/MLL complexes entirely? DPY30 has reported roles in endosomal trafficking that this beat did not examine.
  3. Would the same positioning-versus-catalysis split predict which subunits of SEC and PAF1C are selective? SEC's ELL slot and PAF1C's LEO1 both show essentiality asymmetries that nobody has explained.
Every source behind this entry (4)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    Across 54 DepMap leukemia lines DPY30 correlates with MEN1 at r = 0.446 and survives Holm correction while WDR5, RBBP5 and ASH2L do not, and ASH2L's null result cannot be explained by lack of variance because ASH2L is not pan-essential (21.4% of lines depleted versus DPY30's 31.5%).

    What this evidence cannot show

    Our own computation in 54 lines with wide intervals; the DPY30 and ASH2L confidence intervals overlap slightly at their extremes. No lineage adjustment within the leukemia subset. DepMap release version unverified.

    Evidence that argues against this (1)
    • This project's own deeper re-analysis withdraws the DPY30 result that this evidence item rests on. Against variance-matched nulls drawn from 2,743 usable random protein-coding genes across the same 54 leukemia lines, the Holm-corrected survivors are KMT2A (r 0.783), PSIP1 (0.686), DOT1L (0.681), FLT3 (0.645), MEF2C (0.526) and KAT7 (0.517). DPY30 is explicitly among those that do NOT clear correction, alongside MEIS1 (Holm 0.067), MLLT1 and BCL2. The earlier 16-gene panel gave DPY30 Holm p = 0.015 using an 'exceeds the maximum of the null' criterion that the deep-pool run shows to be unstable by construction, because the null maximum rises mechanically as the pool grows. The DPY30-versus-other-WRAD-subunits contrast is therefore a contrast between one uncorrected near-signal and three nulls, not between a survivor and three failures. Broad Institute DepMap; Open Targets Platform; analysis run by this project 2026
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    DPY30 contributes only modestly to positive heterotypic allosteric regulation of KMT2 methyltransferase activity in vitro while the extensive DPY30-ASH2L interface is critical for correct genome-wide placement of H3K4me2 and H3K4me3, with residual H3K4me2 after DPY30 loss concentrated at highly transcribed genes.

    Structural Analysis of the Ash2L/Dpy-30 Complex Reveals a Heterogeneity in H3K4 Methylation. · Haddad JF, Yang Y, Takahashi YH, Joshi M, Chaudhary N, Woodfin AR, Benyoucef A, Yeung S, Brunzelle JS, Skiniotis G, Brand M, Shilatifard A, Couture JF · Structure 2018
    PMID 30270175DOI
    What this evidence cannot show

    Not performed in leukemia models and does not address MEN1 or KMT2A fusions. In vitro allostery assays are sensitive to complex assembly conditions, which for this complex are strongly temperature- and concentration-dependent.

    Evidence that argues against this (1)
    • This is the decisive objection to the whole WRAD thread and it should be read before any of it. The SET domain of MLL1 -- the catalytic activity that WRAD exists to stimulate -- is dispensable for maintaining haematopoietic stem cells AND for supporting leukemogenesis driven by the MLL-AF9 fusion oncoprotein in mouse. On Mll1 deletion, what is lost at MLL1 target genes is H4K16 acetylation, and inhibiting SIRT1 is sufficient to restore both the acetylation and target-gene expression. The conclusion the authors draw is that recruited MOF activity, not MLL1's intrinsic methyltransferase activity, is what maintains these genes. If H3K4 methylation by MLL1 is not required for MLL-fusion leukemia, then the subunit whose job is to enable that methylation is not a target in this disease, whatever its biochemistry. Mishra BP et al. 2014
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    RBBP5 directly binds H2B-conjugated ubiquitin and orients the MLL complex on the nucleosome, and WDR5 holds the catalytic subunit into the module via the Win motif.

    Structural basis of nucleosome recognition and modification by MLL methyltransferases. · Xue H, Yao T, Cao M, Zhu G, Li Y, Yuan G, Chen Y, Lei M, Huang J · Nature 2019
    PMID 31485071DOI
    What this evidence cannot show

    Cryo-EM of reconstituted catalytic modules on designed nucleosomes; DPY30 is poorly resolved in these maps, so the structural case for its distinct role comes from elsewhere.

    Evidence that argues against this (1)
    • This is the decisive objection to the whole WRAD thread and it should be read before any of it. The SET domain of MLL1 -- the catalytic activity that WRAD exists to stimulate -- is dispensable for maintaining haematopoietic stem cells AND for supporting leukemogenesis driven by the MLL-AF9 fusion oncoprotein in mouse. On Mll1 deletion, what is lost at MLL1 target genes is H4K16 acetylation, and inhibiting SIRT1 is sufficient to restore both the acetylation and target-gene expression. The conclusion the authors draw is that recruited MOF activity, not MLL1's intrinsic methyltransferase activity, is what maintains these genes. If H3K4 methylation by MLL1 is not required for MLL-fusion leukemia, then the subunit whose job is to enable that methylation is not a target in this disease, whatever its biochemistry. Mishra BP et al. 2014
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    A non-active-site surface of the MLL1 SET domain is crucial for the interaction of MLL1 with the RbBP5/Ash2L heterodimer within MLL family core complexes.

    A non-active-site SET domain surface crucial for the interaction of MLL1 and the RbBP5/Ash2L heterodimer within MLL family core complexes. · Shinsky SA, Hu M, Vought VE, Ng SB, Bamshad MJ, Shendure J, Cosgrove MS · Journal of Molecular Biology 2014
    PMID 24680668DOI
    What this evidence cannot show

    In vitro mapping on MLL1 constructs. Note the complication this raises for the hypothesis: the SET domain is absent from KMT2A fusions, so any WRAD contribution to fusion biology must run through the retained wild-type KMT2A allele or through SET1/MLL complexes other than the fusion.

    Evidence that argues against this (1)
    • If MLL1 SET-domain catalysis is dispensable for MLL-AF9 leukemogenesis, then 'ASH2L powers the catalysis' cannot be the reason ASH2L matters in this disease, and the positioning-versus-catalysis dichotomy the hypothesis is built on loses its second arm. The mouse genetics show that H4K16 acetylation by recruited MOF, not MLL1's own methyltransferase output, is what maintains MLL1 target genes, and that SIRT1 inhibition rescues the loss. A version of this hypothesis that survives would have to be about where the complex is placed and what it recruits, with catalysis removed from the model entirely. Mishra BP et al. 2014
Connected to