GSE225028: SUV39H1 and H3K9me3 CUT&RUN in SU-DIPG13 and SU-DIPG36 - the only published human SUV39H1 occupancy map deposited as a readable hg38 bigWig, and the enzyme and its own mark were profiled in the same cells
also called GSE225028 · SUV39H1 CUT&RUN DIPG · Xin 2024 chaetocin DIPG CUT&RUN
A 2024 study measured where the H3K9 silencing enzyme sits on DNA in two brain-tumour cell lines taken from patients, and measured the chemical mark that enzyme writes in the very same cells. The processed files are public, aligned to the standard human genome, and can be read at any gene without doing a new experiment.
The same thing, for a scientist
Four CUT&RUN samples deposited 2023-02-10 as a sub-series of the SuperSeries GSE225029 and published in 2024: SUV39H1 and H3K9me3, each in SU-DIPG13 and SU-DIPG36, both patient-derived H3K27M diffuse midline glioma lines grown as neurospheres in TSM basic medium with B27 (minus vitamin A), hEGF, hFGF, hPDGF-AA, hPDGF-BB and heparin. Antibody CST #8729S for SUV39H1. Reads trimmed, Bowtie2 v2.3.4.1 to **hg38**, MACS2 v2.1.2 narrow and broad, supplementary files deposited as **bigWig**, NovaSeq 6000. For this graph the significance is procedural before it is biological: this is the deposit that converts `SUV39H1 has never been localised anywhere` from never-measured to measured-not-analysed, because a genome-wide hg38 signal track can be read at the KMT2A locus, or anywhere else, with compute alone. Biologically it is the rarer thing - **a writer and its own product profiled in the same nuclei**, which is what is needed to ask whether SUV39H1 residency and H3K9me3 deposition are actually coextensive or whether the enzyme is found where the mark is not.
Why a writer-and-mark pair in the same cells is the valuable part
Almost every H3K9me3 argument in this graph is an argument about a mark, inferred back to an enzyme nobody localised. GSE225028 has both sides in the same nuclei, so the inference can be checked rather than assumed 1. That question - does SUV39H1 sit only where H3K9me3 is - has a direct bearing on the graph's rearrangement-brake reasoning, because a brake that depends on the mark and a brake that depends on the enzyme predict different things when the enzyme is inhibited and the mark decays on its own schedule.
The dimensions the PI asks for
Cell type: patient-derived diffuse midline glioma, H3K27M-mutant, grown as neurospheres - a neural progenitor-like state, not a haematopoietic one. Cell state: stem-like/self-renewing; the paper's own framing is that chaetocin-mediated SUV39H1 inhibition targets stemness networks, so the profiled state is the one the drug is aimed at. Subcellular and subnuclear localization: the assay reports chromatin-bound SUV39H1 only; nothing in the deposit speaks to a soluble pool, and no subnuclear imaging accompanies it. Epigenetics: H3K9me3 CUT&RUN in the same two lines, on the same pipeline, is the matched readout - and the H3K27M background means the H3K27me3 landscape in these cells is globally redistributed, which is a confound for any heterochromatin generalisation drawn from them 2.
What this deposit cannot be asked
There is no SUV39H1-knockout or IgG-only control track in the deposit, so a peak in the SUV39H1 track is bounded by the antibody's specificity and not by a genetic control. The CST #8729 clone is recorded here because rule 74 requires the reagent identifier rather than the label, and because a SUV39H1 antibody with cross-reactivity to SUV39H2 would produce exactly this data with a different meaning. Two lines is n=2 systems, both of the same disease, both H3K27M - so a site present in both is replicated within a genotype and not across one.
| design | CUT&RUN for SUV39H1 (CST #8729S) and for H3K9me3, in two patient-derived H3K27M diffuse midline glioma lines (SU-DIPG13, SU-DIPG36) grown as neurospheres |
|---|---|
| readout | genome-wide chromatin occupancy; deposited as per-sample bigWig on hg38 |
| controls | no SUV39H1-knockout track and no IgG track in the deposit; H3K9me3 in the same cells serves as a biological co-readout, not as a specificity control |
| cost tier | low |
| timeline | already run and deposited; reading the tracks at any locus is compute only |
| accession | GSE225028 (sub-series of SuperSeries GSE225029) |
| pmid | 38011799 |
| genome build | hg38, stated per sample in !Sample_data_processing |
| pipeline | FastQC v0.11.7, Bowtie2 v2.3.4.1, MACS2 v2.1.2 narrow and broad |
| instrument | Illumina NovaSeq 6000 |
| antibody clone | Cell Signaling Technology #8729S (SUV39H1) |
| submission date | 2023-02-10 |
| cell type | patient-derived diffuse intrinsic pontine / diffuse midline glioma, H3K27M-mutant, neurosphere culture |
| cell state | stem-like, self-renewing; the state the paper's drug arm targets |
| localization | chromatin-bound fraction only; no soluble-pool or imaging data in the deposit |
| subnuclear localization | not measured in this deposit |
| epigenetic readout | SUV39H1 occupancy and H3K9me3 occupancy, same cells, same pipeline |
| shown for fusion | NO - nobody has looked in any KMT2A fusion system. This is a glioma deposit; its relevance here is that it is a readable human SUV39H1 track, not that it is a leukaemia experiment. |
| why it is in this graph | it is the deposit that converts the SUV39H1 half of absence-no-h3k9-writer-or-reader-localised-at-the-kmt2a-locus from never-measured to measured-not-analysed |
samples
[ "GSM7037529", "GSM7037530", "GSM7037531 (SU-DIPG13, SUV39H1)", "GSM7037532 (SU-DIPG36, SUV39H1)" ]
- The pharmacological arm of the source paper uses chaetocin, which is not a selective SUV39H1 inhibitor; any therapeutic inference that travels from this deposit into a KMT2A programme inherits that.
- SUV39H1 loss is being engineered deliberately into CAR T products to improve persistence, and several of the GEO samples this sweep returned are exactly those constructs. A SUV39H1-directed agent in leukaemia would therefore be pushing on a target that adoptive-cell programmes are removing on purpose, in the same patients.
- Do SUV39H1 and H3K9me3 coincide in these cells, and what fraction of SUV39H1 sites carry no H3K9me3? A writer found off its own mark is either a scaffolding role or an antibody artefact, and the same deposit can distinguish the two only if a knockout control is generated. ANSWERED IN CYCLE 66 by ana-q114-the-h3k9me3-writer-and-its-own-mark-barely-coincide-in-the-same-nuclei: over 59,821 primary-assembly CTCF-summit windows read from these four bigWigs, Spearman rho between SUV39H1 and H3K9me3 is 0.148 (SU-DIPG13) and 0.078 (SU-DIPG36) - essentially the same as over 200,000 random tiles (0.159, 0.085). 25.4% and 39.1% of the top 1% of SUV39H1 sites carry H3K9me3 BELOW the 50th percentile of anchors; the median mark percentile at those sites is 76.6 and 61.1. The writer IS substantially off its own mark. This entry's own diagnosis stands: scaffolding versus antibody artefact cannot be separated, because the deposit has no IgG and no knockout track.
- Is there signal over the KMT2A gene body in either line? Deliberately not run here: a concurrent beat owns signal-level work at the KMT2A exon-11 anchor and the two analyses must not collide. ANSWERED IN CYCLE 66, by the beat that owned the collision: ana-q114-suv39h1-is-not-at-the-kmt2a-locus-in-either-patient-derived-glioma-line. NO. Over the 90,377 bp gene body SUV39H1 reads 0.9381 and 1.1109 - 0.94x and 1.11x the genome-wide random-tile mean, and the 39.7th and 49.7th percentile of width-matched CTCF-anchor windows. All seven KMT2A intervals are NOT OCCUPIED in both lines, and the 147 bp exon-11 anchor reads an EXPLICIT ZERO in SU-DIPG13.
- Is the SU-DIPG13 versus SU-DIPG36 difference in SUV39H1 distribution larger or smaller than the difference between an H3K27M line and a non-K27M line? Without a non-K27M comparator in the deposit the H3K27M confound cannot be sized.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
SUV39H1 and H3K9me3 CUT&RUN were run in SU-DIPG13 and SU-DIPG36 and deposited as hg38 bigWigs.
Chaetocin-mediated SUV39H1 inhibition targets stemness and oncogenic networks of diffuse midline gliomas and synergizes with ONC201 · Xin D, Liao Y, et al. · Neuro-Oncology 2024What this evidence cannot show
The design, build, antibody and file format are read from the GEO SOFT records of GSM7037531 and GSM7037532 and from the series' own overall_design string, fetched 2026-09-22. The paper's figures were not opened; nothing here is a claim about what the paper concluded.
Evidence that argues against this (1)
- Chaetocin is a thiodioxopiperazine with thiol-reactive, redox-cycling activity and is a poor selectivity tool for SUV39H1; a body of work argues its cellular effects are largely oxidative rather than methyltransferase-specific. That weakens the paper's pharmacological arm, though not the occupancy deposit this entry is about. Cherblanc FL et al. 2013
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
SU-DIPG13 and SU-DIPG36 carry the H3K27M mutation, whose global effect on the repressive chromatin landscape makes them an atypical background for a heterochromatin generalisation.
Reduced H3K27me3 and DNA hypomethylation are major drivers of gene expression in K27M mutant pediatric high-grade gliomas · Bender S, Tang Y, Lindroth AM, et al. · Cancer Cell 2013What this evidence cannot show
This citation establishes the H3K27M chromatin phenotype in general, not the genotype of these two specific lines; the line identity is taken from the GEO sample characteristics, which name them as patient-derived DIPG cells.
Evidence that argues against this (1)
- The Bender result is about H3K27me3, and it does not establish how H3K9me3 behaves in these cells - so the size and even the direction of the confound for an H3K9 readout is not settled by it. A H3F3A-mutant paediatric GBM series reports the global trimethylation code as altered more broadly, which argues the confound is real but does not quantify it for H3K9me3 in a K27M line. Only two records in PubMed pair K27M with H3K9me3 at title/abstract level and neither is a DIPG occupancy study, so this is a genuinely open point rather than a settled objection. Pathak P et al. 2015
the deposit that carries the overturn: hg38 bigWigs, readable at any locus with compute alone
the only published human SUV39H1 occupancy map with its own mark profiled in the same nuclei
first reading of the deposit that entry describes
answers this entry's first open question: rho 0.148 and 0.078
this deposit's four bigWigs carry two different chrom tables, 455 and 195