Oncofusion

start / graph / exp-q114-a-suv39h1-cutandrun-in-a-cd34-progenitor-and-a-kmt2a-rearranged-line

Experimentstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q114-suv39h1-at-the-locus

The one experiment that would close the SUV39H1 arm of this graph's H3K9 argument: CST #8729S CUT&RUN in a CD34+ progenitor and in SEM or RS4;11, with an IgG arm and a SUV39H1-degron arm that neither existing deposit has - because after this cycle the enzyme has been measured in exactly two human cell types and neither is a blood cell

also called the CD34+ SUV39H1 CUT&RUN · close the SUV39H1 arm

Everything public about where this enzyme sits on human chromosomes comes from two cell types: a children's brain tumour and a breast cancer line. One experiment in blood-forming cells and one in a leukaemia line with the relevant gene fusion would settle the question this project keeps having to answer by substitution.

The same thing, for a scientist

**The state of the evidence after cycle 66, which is what makes this experiment the priority rather than a nice-to-have.** Human SUV39H1 occupancy exists in **two** cell types in all public archives: SU-DIPG13/SU-DIPG36 (GSE225028, CUT&RUN, CST #8729S) and MCF-7 (GSE139196, ChIP-seq, no clone stated). ENCODE holds nothing under any label, by three controlled routes. ChIP-Atlas holds the MCF-7 four re-indexed and nothing else [E1][E2]. **Neither system is haematopoietic, neither is *KMT2A*-rearranged, both are resting and unirradiated, and neither deposit contains an IgG or a knockout control.** **Q114 read both at the *KMT2A* locus and found the enzyme absent in the DIPG lines** [E3]. That is a real result and it bounds two glioma lines. It cannot be carried into infant B-ALL, and last cycle showed exactly why at this very element: BRCA1 sits at the **93.7th and 92.3rd** percentile of same-cell CTCF anchors at the exon-11 anchor in HeLa-S3 and H1, and at the **34.0th and 45.1st** in the only two blood lines, where ENCODE's 300,000-peak pre-IDR lists carry no BRCA1 peak at all - **occupancy at this anchor is cell-type-restricted**. **Design.** CUT&RUN, **CST #8729S** (the clone that already has a deposited human track, so the new data joins the old), in three systems: (i) CD34+ cells from cord blood or mobilised marrow, the compartment this disease's putative origin sits in; (ii) **SEM** and **RS4;11**, both *KMT2A*::AFF1 and both already in this graph's normalisation batch from GSE63664; (iii) **REH** as a *KMT2A*-germline B-ALL comparator, which is the control arm that GSE63664 experiment already uses. **Two controls that decide whether the answer means anything, and that neither existing deposit has:** - an **IgG** arm, because with no IgG a low signal and a low antibody accessibility are the same number; - a **SUV39H1 degron or knockout** arm in at least one line. Q114 measured Spearman 0.148 and 0.078 between SUV39H1 and H3K9me3 in the same nuclei, so a quarter to two-fifths of the enzyme's strongest sites carry no more of its own mark than a median CTCF anchor. **That is either non-catalytic residency or antibody cross-reactivity, and only a knockout separates them** - the corpus entry that catalogued the deposit said so first and it is right. **Readout and pre-registered analysis.** The seven stored *KMT2A* intervals, scored against same-cell-type CTCF anchors from a CTCF ChIP or CUT&RUN run in the same cells - which also supplies the C4 instrument control that both Q114 arms lacked, and which is the reason to run CTCF alongside rather than borrow an anchor set from another cell type. **Add a damage arm if budget allows.** `hyp-the-writer-reaches-the-kmt2a-anchor-only-through-ctcf-and-only-after-the-break` predicts *damage-conditional* recruitment, which a resting measurement cannot touch. Etoposide or an AsiSI/CRISPR cut at the element with a short time course would test it, and **no damage-state occupancy measurement exists at this locus in any cell type at all**.

Go deeper
Why this is cheap relative to what it settles

One antibody, one assay, three to five cell inputs, plus a matched CTCF arm. CUT&RUN needs 10^5 cells or fewer, which is why it is the assay of choice for primary CD34+ material where a ChIP would not be feasible.

What it closes: the entire SUV39H1 arm of this graph's H3K9 argument, which currently rests on (a) a literature absence bounded by titles and abstracts, (b) a menin-delivery model from four endocrine systems, (c) an ENCODE zero, and (d) two non-haematopoietic occupancy reads. exp-suv39h1-chip-seq-in-mcf7-with-and-without-estradiol already named this as its own third open question - would that single experiment close the whole SUV39H1 arm of this graph's H3K9 argument? - and the answer is that it would close the occupancy half of it.

What it would NOT close
  • Causality. Occupancy is not deposition. The Q114 writer-versus-mark correlation says those are different measurements even in the same nucleus.
  • The menin link. interface-menin-suv39h1-svid-360-445 brackets a menin surface that binds SUV39H1, and menin removal from a promoter removes SUV39H1 from it in an endocrine system. Whether that route operates in a KMT2A-rearranged blood cell needs a menin-inhibitor arm, not a resting occupancy map.
  • The fusion. Nothing in this design measures the fusion protein. shown_for_fusion stays NO until someone runs SUV39H1 occupancy in a line where the fusion is present and a matched line where it is not.
Facts and figures
designCUT&RUN for SUV39H1 with Cell Signaling #8729S in (i) CD34+ cord-blood or mobilised-marrow progenitors, (ii) SEM and RS4;11 (KMT2A::AFF1), (iii) REH (KMT2A-germline B-ALL comparator); matched CTCF CUT&RUN in the same cells to supply a same-cell-type class-matched anchor null and the C4 instrument control; optional damage arm with etoposide or an AsiSI/CRISPR cut at the exon-11 element plus a short time course
readoutgenome-wide occupancy; primary statistic = mean signal over the 147 bp exon-11 anchor as a percentile of same-cell-type CTCF summit windows, with the six other stored KMT2A intervals as the secondary readout
controlsIgG arm (absent from both existing deposits); a SUV39H1 degron or knockout arm in at least one line (absent from both existing deposits); same-cell-type CTCF for the class-matched null and for C4; housekeeping-promoter and gene-desert internal controls as in the Q114 pre-registration
cost tiermed
timelineweeks; CUT&RUN needs <=1e5 cells, which is what makes the primary CD34+ arm feasible where a ChIP would not be
cell typeCD34+ haematopoietic progenitor; SEM; RS4;11; REH
cell stateresting, plus an optional etoposide or induced-break damage state
lineagehuman haematopoietic - the lineage no existing SUV39H1 occupancy measurement reaches
shown for fusionNO, and that is the point: SEM and RS4;11 versus REH is the first design in which SUV39H1 occupancy would be measured in a KMT2A::AFF1 cell and a matched germline comparator
antibody cloneCell Signaling Technology #8729S - chosen because it is the clone with an existing deposited human track, so the new data can be compared to GSE225028 rather than only to itself
pre registration requiredthe null (same-cell-type CTCF anchors), the threshold and the prior must be written before the tracks are read; the Q114 pre-registration is the template, and its three failed controls (C1 degenerate, C1b failed, C4 unrunnable) are the three things to fix in the next version
what it closesthe occupancy half of the SUV39H1 arm of this graph's H3K9 argument
what it does not close
[
  "causality - occupancy is not deposition",
  "the menin-delivery route, which needs a menin-inhibitor arm",
  "any statement about the fusion protein itself"
]
How this could bite us back
  • A positive result here would immediately be read as a rationale for a SUV39H1 inhibitor in KMT2A-rearranged leukaemia, and the only SUV39H1 agent with clinical traction is chaetocin - a promiscuous thiodioxopiperazine whose phenotypes cannot be attributed to SUV39H1 catalysis. The experiment should be designed and reported so that it cannot be used that way.
  • SUV39H1 is at Xp11.23 and is the dominant H3K9 methyltransferase in the haematopoietic system, so an inhibitor's on-target toxicity would fall on normal progenitors - and X copy number and X-inactivation status vary line to line and patient to patient, which nothing in this literature reports. A CD34+ arm should record donor sex.
  • A clean negative here would close the arm, and closing an arm is how this project has previously suppressed work for seven cycles on one stale number. Any negative should be recorded with the identities and the per-interval numbers persisted, not as a summary figure.
Still open
  1. Is CST #8729S validated in a knockout for CUT&RUN, or only for western blot? The clone's existing human occupancy track has no knockout control behind it, so the proposed degron arm is also a validation of the reagent this graph would then depend on.
  2. Should the comparator be REH or an isogenic KMT2A::AFF1 knock-in? REH differs from SEM and RS4;11 in far more than the fusion, so a fusion-attributable difference in SUV39H1 occupancy could not be claimed from a three-line panel. An isogenic pair is the stronger design and is much more expensive.
  3. Would a CUT&RUN in primary infant KMT2A::AFF1 blasts be possible at all? The graph records that no menin-inhibitor occupancy dataset exists in an infant-derived KMT2A::AFF1 line, which suggests primary infant material is the binding constraint on every occupancy question in this disease.
Every source behind this entry (3)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    SUV39H1 has no ENCODE target record and no ENCODE experiment under any label, by three routes queried 2026-09-22 with a positive and a negative control through the identical URL template.

    What this evidence cannot show

    An archive absence is dated and archive-scoped. It does not exclude an experiment filed under an unrelated target label, and it says nothing about unpublished data.

    Evidence that argues against this (1)
    • The registry row on this trap notes that ReMap returns 401 for SUV39H1 - an access failure, not an absence - and this beat did not retry it. GTRD was not queried. So 'two systems' is a floor from the archives actually reached. source
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    The only human SUV39H1 occupancy deposits are GSE225028 (CUT&RUN, CST #8729S, two DIPG lines) and GSE139196 (four ChIP-seq, MCF-7, no clone stated), and neither contains an IgG or a knockout control.

    GSE225028 and GSE139196 per-sample SOFT records · Xin DE et al.; Yang F et al. · NCBI GEO 2023
    What this evidence cannot show

    Read 2026-09-22 from acc.cgi per-sample brief records for both series (4 and 77 samples respectively). 'No IgG identified' is a statement about the sample titles and characteristics fields; an unlabelled control sample cannot be ruled out.

    Evidence that argues against this (1)
    • GSE139196 does contain an shKDM4B+shKDM4C perturbation arm for H3K9me3 and two HP1-alpha mutant Bio-ChIPs, so it is not control-free in every respect - it is control-free for SUV39H1 specifically. source
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    Q114 read both deposits at the KMT2A locus against a pre-registered class-matched null and found SUV39H1 NOT OCCUPIED at all seven intervals in both DIPG lines.

    Q114 pre-registration and results (analyses/q114-suv39h1-at-the-kmt2a-locus) · Q114-suv39h1-at-the-locus · oncofusion-knowledge-graph 2026
    What this evidence cannot show

    The DIPG arm's heterochromatin positive control FAILED at four window scales, so the deposit's sensitivity at a 147 bp element is not established; and the arm was declared C4-blind in advance because no CTCF experiment exists in any DIPG line.

    Evidence that argues against this (1)
    • A negative in a deposit whose positive control fails is exactly the result a poor library would produce. This is the strongest single reason the experiment proposed here is needed rather than optional. source
Connected to