ATRX - DAXX's partner and the other arm of H3.3 deposition; in this disease its only signal is in the four lymphoid cell lines
also called ATRX · XH2 · RAD54L · XNP · alpha-thalassemia/mental retardation syndrome X-linked · P46100
ATRX works with DAXX to slot the replacement histone into the silent, tightly packed parts of the genome - the opposite territory from the genes a KMT2A fusion forces open. Which of those two territories the fusion actually touches is the question this whole beat exists to ask.
The same thing, for a scientist
ATRX (P46100, 2,492 aa) is a SWI/SNF-family chromatin remodeller that partners DAXX to deposit H3.3 at heterochromatin. The founding genome-wide result is negative as much as positive: **Atrx is required for HIRA-independent H3.3 localisation at telomeres and for repression of telomeric RNA, and Atrx/Daxx associate with H3.3 independently of Hira** [E1]. The heterochromatin arm and the gene-body arm are therefore biochemically separable, which is the premise of this beat's central question. **The KMT2A intersection is empty on title check.** `ATRX AND (MLL OR KMT2A)` returns 9 records and all 9 are gene-panel or exome co-mentions - JMML exome sequencing, an MDS prognostic score, neurodevelopmental episignatures, an ovarian line, a neuroendocrine fusion survey, an EGFR anti-PD-1 signature, a cognition review, plasmablastic lymphoma, and exome pool-seq [E3]. Not one studies ATRX in a KMT2A-rearranged cell. **The DepMap result is the interesting part, and it is a lineage result.** ATRX's overall difference is -0.115 at a pan-essentiality of -0.204, which is unremarkable. Split by lineage it is **-0.0003 in the 7 myeloid index lines and -0.321 in the 4 lymphoid ones** [E4]. The entire signal lives in the half of the arm that matches this graph's index disease - and rests on four cell lines.
Which arm would a KMT2A fusion touch?
The beat's question, stated precisely: a KMT2A fusion's proposed mechanism is forced transcriptional elongation through gene bodies. Forced elongation evicts and re-deposits histones in gene bodies, which is HIRA territory 1. On mechanism alone the prediction is that the fusion perturbs the gene-body arm and leaves the heterochromatin arm alone.
Three things complicate that, and all three point at ATRX/DAXX anyway:
- The dependency data point the other way. HIRA has no lymphoid selectivity (+0.032) and ATRX and DAXX both do (-0.321 and -0.601) 4. If the screens mean anything, the arm the fusion needs is the heterochromatin arm, not the gene-body arm.
- The signal is in four lines. With n = 4, the lymphoid halves are the least reliable numbers in the beat and this project's sibling shortlist of larger differences did not survive a permutation null.
- A differentiation-blocked leukaemia has a heterochromatin problem. H3.3-null haematopoietic progenitors lose H3K9me3 at endogenous retroviral repeats and open IRF binding sites (
protein-histone-h3-3E4) - a plausible route by which the ATRX/DAXX arm could matter in a blast without the fusion touching it directly.
All three are reasons to look, and none is evidence. The entry is written so a future cycle can kill the claim cheaply.
Where ATRX does have blood literature, and why it is a comparator
ATRX has a real haematological literature, and it is not about H3.3 deposition. Acquired ATRX point mutation or splicing defect causes acquired alpha-thalassaemia myelodysplastic syndrome (ATMDS), a recognised entity - although the same study that names it also confirms it is rare, 0% in an unselected cohort of 80 low-risk MDS 5. That is a real, patient-level, blood-marrow phenotype for this gene - in a different disease, through a different mechanism (alpha-globin locus regulation), in adults.
Naming it matters because it is the kind of literature that makes an absence look filled. A reader who searched ATRX AND leukemia would find records and could conclude the gene is studied in blood. It is - as a globin-regulation and MDS gene. It is not studied as an H3.3 chaperone in any leukaemia.
| uniprot | P46100 |
|---|---|
| gene symbol | ATRX |
| length AA | 2492 |
| partner | gene-daxx |
| deposits | protein-histone-h3-3 |
| locus class | telomeres, pericentric heterochromatin, endogenous retroviral repeats |
| kmt2a literature records title checked | 0 |
| alias hazard | RAD54L is a curated alias symbol of ATRX and also a distinct gene (rule 63) |
depmap — the numbers behind the chart
{
"index_mean": -0.4024,
"other_leukaemia_mean": -0.2872,
"difference": -0.1152,
"pan_essentiality": -0.2044,
"myeloid_difference": -0.0003,
"lymphoid_difference": -0.3214,
"n_index": 11,
"n_index_myeloid": 7,
"n_index_lymphoid": 4
}depmap kmt2a selectivity
{
"chronos_mean_kmt2a_rearranged": -0.4024,
"chronos_mean_other_leukaemia": -0.2872,
"difference": -0.1152,
"pan_essentiality_all_lines": -0.2044,
"selectivity_rank": 853,
"rank_denominator": 17728,
"provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
"how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}- ATRX is a 2,492-residue chromatin remodeller required for telomere and repeat silencing in every cell; loss of ATRX function is the classic route to alternative lengthening of telomeres, so inhibiting it risks supplying a telomere-maintenance mechanism to the tumour.
- Germline ATRX loss causes an X-linked intellectual disability and alpha-thalassaemia syndrome, and acquired ATRX mutation in adults presents as acquired alpha-thalassaemia - two independent indications that partial loss of this gene has haematological consequences in humans.
- The only signal for ATRX in this disease is in four cell lines in one half of one screen, which is the evidential profile of a claim that will not replicate; building anything on it is the failure mode this graph has already paid for once with DAXX.
- Is the lymphoid-restricted ATRX/DAXX dependency real? The cheapest test is a permutation null on the same 11-versus-43 comparison plus the lineage split, which this beat did not compute and which killed the sibling shortlist.
- Does KMT2A::AFF1 occupancy overlap ATRX/DAXX-dependent H3.3 domains, or only HIRA-dependent gene bodies? Nobody has done the overlap, and it is the question that decides whether this arm belongs in the graph at all.
- Does the newer model - DAXX delivering pre-methylated H3.3K9me3 rather than maintaining the mark after deposition - change what an ATRX loss would do in a differentiation-blocked blast?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Atrx and Daxx associate with H3.3 in a Hira-independent manner; Atrx is required for Hira-independent localisation of H3.3 at telomeres and for repression of telomeric RNA, while Hira is required for H3.3 enrichment at active and repressed genes but not at telomeres.
Distinct factors control histone variant H3.3 localization at specific genomic regions. · Goldberg AD, Banaszynski LA, Noh KM, et al. · Cell 2010What this evidence cannot show
Abstract read from efetch 2026-09-20; full text not read. Mouse ES and neural precursor cells with a zinc-finger-nuclease-tagged H3.3 allele. Telomeres are the locus class where the ATRX requirement is demonstrated; extension to pericentric heterochromatin and endogenous retroviruses comes from other studies not read for this entry.
Evidence that argues against this (1)
- The division of labour is not fixed across cell types: in fetal liver haematopoiesis HIRA is dispensable and ASF1B determines H3.3 enrichment, so 'ATRX owns heterochromatin, HIRA owns gene bodies' is an ES-cell map that has never been drawn in a blood cell. Liu J et al. 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DAXX provides a distinct function within the histone chaperone network by recruiting histone methyltransferases to catalyse H3K9me3 on new histone H3.3-H4 before deposition onto DNA, giving a molecular mechanism for de novo H3K9me3 deposition and heterochromatin assembly; the same exploratory interactomics defines the human H3-H4 chaperone network and predicts an ASF1-SPT2 co-chaperone complex.
DAXX adds a de novo H3.3K9me3 deposition pathway to the histone chaperone network. · Carraro M, Hendriks IA, Hammond CM, et al. · Molecular Cell 2023What this evidence cannot show
Abstract read from efetch 2026-09-20; full text not read. Human cell lines, interactomics plus targeted biochemistry; the ASF1-SPT2 complex is a structure PREDICTION, not a solved structure, which the abstract states. The de novo methylation claim is about DAXX rather than ATRX, and this entry cites it because ATRX's partner's mechanism changed - it is not evidence about ATRX itself.
Evidence that argues against this (1)
- This reframes the ATRX/DAXX arm: the older model has the complex MAINTAINING H3K9me3 at heterochromatin, this one has DAXX DELIVERING pre-methylated histone. Both are in the literature and this graph's gene-daxx entry currently carries only the older framing. Goldberg AD et al. 2010
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
As of 2026-09-20, PubMed returns 9 records for 'ATRX AND (MLL OR KMT2A)' and all 9 are off-topic on title check: a JMML whole-exome epigenetic-regulator survey, an MDS molecular-evolution prognostic score, an episignature evaluation for neurodevelopmental disorders, a chromosome-instability analysis of ovarian lines, a fusion survey of ACTH-secreting neuroendocrine neoplasms, an anti-PD-1 mutation signature in EGFR lung cancer, a cognitive-genetics review, a plasmablastic lymphoma series, and an exome pool-seq method paper. None studies ATRX in a KMT2A-rearranged cell.
NCBI E-utilities esearch over PubMed for ATRX against KMT2A and MLL with title checking of all returned records, run 2026-09-20 · US National Library of Medicine 2026What this evidence cannot show
Query translation captured and contained no [UID] term. All 9 titles were fetched in full from efetch XML using itertext(), not read from a truncated listing. **Bounded by its symbols**: ATRX's aliases XH2, XNP and RAD54L were not separately crossed with MLL/KMT2A, and RAD54L is a distinct gene so that alias is unusable. A KMT2A-fusion paper naming ATRX only in a supplementary table would not be caught.
Evidence that argues against this (1)
- A symbol-pair zero is weakest for exactly the large-scale studies that would be most likely to contain the observation, and this project has documented a 2024 interactome map that recovers AFF1 without naming it anywhere in the body text. Carraro M et al. 2023
- E4weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset
In DepMap CRISPR gene effect, ATRX's mean effect in 11 KMT2A-rearranged leukaemia lines is -0.402 against -0.287 in 43 other leukaemia lines, a difference of -0.115 at a pan-essentiality of -0.204 across 1,208 lines; split by lineage the difference is -0.0003 in the 7 myeloid index lines and -0.321 in the 4 lymphoid index lines. DAXX shows the same pattern (-0.117 myeloid, -0.601 lymphoid) while HIRA reverses sign (-0.147 myeloid, +0.032 lymphoid).
Is any histone chaperone or variant-exchange gene selectively required in KMT2A-rearranged leukaemia lines, and does the answer survive a lineage split? · Oncofusion knowledge graph (internal analysis) 2026What this evidence cannot show
Run 2026-09-20. **Weak, and deliberately so.** The lymphoid half of the index arm is **four cell lines**. No permutation null was computed. The overall difference (-0.115) is below this project's -0.15 convention and only the lineage-split half clears it, which is the exact shape of a subgroup finding that does not replicate. This project's sibling triage produced a shortlist with larger differences that a 50,000-label permutation null subsequently killed, DAXX included at p = 0.0128 and failing Bonferroni. DepMap contains **no infant KMT2A::AFF1 line**.
Evidence that argues against this (1)
- The DAXX entry in this graph records that a 50,000-label permutation null killed its own dependency claim, and DAXX's difference (-0.305) is nearly three times ATRX's. A weaker number from the same instrument on the same lines cannot be stronger evidence. Oncofusion knowledge graph (internal analysis) 2026
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Acquired alpha-thalassaemia myelodysplastic syndrome (ATMDS) is an acquired syndrome characterised by a somatic point mutation or splicing defect in ATRX in patients with myeloid disorders, primarily MDS; in a study focused on MDS with unexplained microcytosis the data confirm a LOW frequency of ATRX mutations - 0% in an unselected clinical-trial cohort of 80 low-risk MDS and 0.2-0.8% in a multicentre series.
Incidence of ATRX mutations in myelodysplastic syndromes, the value of microcytosis. · Herbaux C, Duployez N, Badens C, et al. · American Journal of Hematology 2015What this evidence cannot show
Abstract read from efetch 2026-09-20; full text not read. Adult MDS, not paediatric and not KMT2A-rearranged. **Note the direction of the result: this paper CONFIRMS that ATRX mutation is rare in MDS** (0% of 80 low-risk cases), so it establishes the entity ATMDS and simultaneously bounds how common it is. Cited here only to show that ATRX has a genuine haematological literature which is NOT about H3.3 deposition, so a future reader does not mistake the presence of blood papers for coverage of the question this beat asks.
Evidence that argues against this (1)
- The ATMDS phenotype works through alpha-globin locus regulation, not through H3.3 chaperoning, and the reviews that established the entity frame it as an acquired haemoglobinopathy. Citing a real blood literature for a gene does not close the gap this beat identified and could be used to make it look closed. Steensma DP et al. 2005
- E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
ATRX is UniProt P46100 (ATRX_HUMAN), chromatin remodeler ATRX / transcriptional regulator ATRX, 2,492 amino acids, with alias symbols RAD54L and XH2.
UniProtKB P46100 (ATRX_HUMAN), Chromatin remodeler ATRX, 2492 aa · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026What this evidence cannot show
Fetched 2026-09-20 with `gene_primary` requested and read back per rule 62: exactly one reviewed human record with primary gene name ATRX. **The alias list contains RAD54L, which is a different gene with its own record** - rule 63's trap exactly. Any alias-expanded literature search on ATRX must exclude RAD54L hits, and this entry's literature counts were run on the primary symbol only for that reason.
Evidence that argues against this (1)
- Because RAD54L is a curated alias of ATRX and also a real separate gene, a tool matching alias fields can resolve either string to the wrong record. The accession is asserted here so a length check (2,492 aa) can catch a swap. UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
ATRX/DAXX deposit H3.3 at heterochromatin independently of HIRA
- recruits →Histone H3.3 - one protein, two genes, and the reason this graph should not file it as gene-h3f3a
required for HIRA-independent H3.3 localisation at telomeres
ATRX-DAXX maintain H3K9me3 at heterochromatin alongside H3.3 deposition
ATRX is required for HIRA-independent H3.3 localisation at telomeres
DEK loss displaces ATRX from telomeres and redistributes H3.3
ATRX forms a complex with SETDB1, TRIM28 and ZNF274 co-occupying the 3' exons of zinc-finger genes, the one uncontroversial genomic address of this writer
ATRX/DAXX inhibit the heterochromatin-driven PML-telomere colocalisation that tethered HP1-alpha nucleates
the ATRX half of the lymphoid-restricted claim does not survive: p 0.101 at donor level and 0.486 against a B-ALL comparator
ATRX likewise absent from all 960 rows
ATRX loss is the canonical human route to ALT, documented in 61% of panNET; never measured in this disease
targets the telomeric arm of H3.3 deposition that ATRX, not HIRA, controls
no human ChEMBL target record at all; the only record is murine with two off-target phospho rows