SIRT1 - not a fusion partner, the least required gene in KMT2A-rearranged lines, and the one target here where the direction of the drug is genuinely disputed
also called SIRT1 · SIR2L1 · NAD-dependent protein deacetylase sirtuin-1 · sirtuin 1
SIRT1 is an enzyme that switches genes off, and it is not something a KMT2A fusion recruits. Whether the useful drug here turns SIRT1 on or off is genuinely disputed, and nobody has run the two directions side by side in KMT2A-rearranged leukaemia.
The same thing, for a scientist
SIRT1 (Q96EB6, 747 aa) is an NAD-dependent class III protein deacetylase. Its place in this interactome is narrower than it is usually presented: Chen et al. 2015 show that DOT1L **inhibits SIRT1-mediated epigenetic silencing** to maintain leukemic gene expression in MLL-rearranged leukemia, i.e. SIRT1 is a **repressor acting downstream of DOT1L inhibition**, not a protein recruited by an MLL fusion. **No verified paper shows a direct physical interaction between SIRT1 and any MLL/KMT2A fusion protein**, and none is claimed here. DepMap makes SIRT1 the least-required gene in the KMT2A-rearranged panel: mean Chronos +0.035 across 1,208 lines and **+0.116 across the 10 KMT2A-rearranged lines**, the most positive KMT2A-rearranged value of any gene in this dossier. The therapeutically interesting consequence of Chen et al. is that SIRT1 **activators** combine with DOT1L inhibitors, and **SRT-2104 (CHEMBL4297436) is a phase-2 SIRT1 activator that already exists**. **That direction is disputed, and this entry no longer asserts it.** SIRT1/2 inhibition with tenovin-6 activates p53, kills primary ALL cells from 41 children and sensitises them to etoposide and cytarabine [counter-evidence to E2], and SIRT1 knockout suppresses BCR-ABL leukaemogenesis [counter-evidence to E2]. No activator-versus-inhibitor comparison has been run in a KMT2A-rearranged model, and no compound in this literature is SIRT1-selective enough to settle it.
Tell me more about this picture
PDB 4ZZJ, X-ray 2.74 A: SIRT1 residues 183-505 with an activator and a p53 substrate peptide.
Tell me more about this picture
This is the most informative entry for the agonist direction, which PMID 25822366 predicts. The direction is disputed: SIRT1/2 inhibition kills primary paediatric ALL cells (PMID 25884180) and SIRT1 knockout suppresses BCR-ABL leukaemogenesis (PMID 22207735).
Tell me more about this picture
For biasing the conformational equilibrium, the companion set is 4KXQ (1.85 A, closed) and 4IG9 (2.64 A, open) — the catalytic domain with its C-terminal regulatory segment in both end states.
Tell me more about this picture
The best-characterised SIRT1 co-crystal, 4I5I at 2.50 A, contains an EX-527 analogue - an INHIBITOR. Previously flagged here as a trap; as of 2026-09-19 it is the comparator arm the deciding experiment needs, because the direction is not settled.
Tell me more about this picture
DepMap Chronos means from the Open Targets API, computed by us from per-line values; n=10 KMT2A-rearranged lines, so differences under 0.15 are noise.
Tell me more about this picture
The sign agrees with the DOT1L-epistasis mechanism (PMID 25822366): if SIRT1 silences the leukaemic programme, its loss should be neutral-to-favourable. It disagrees with the reports in which SIRT1 is an oncogenic survival pathway (PMID 22207735, PMID 25884180) - the DepMap sign does not adjudicate, because an unperturbed screen is the wrong assay for either model.
Tell me more about this picture
On the DOT1L-epistasis model a SIRT1 inhibitor should antagonise a DOT1L inhibitor. That prediction has never been tested in a KMT2A-rearranged model and the opposite direction has positive results in paediatric ALL (PMID 25884180).
This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
What SIRT1 is not
This section exists because the mis-statement is common and it inverts the whole therapeutic logic.
SIRT1 is not a KMT2A-fusion-recruited cofactor. Chen et al. 2015 report that DOT1L inhibits SIRT1-mediated epigenetic silencing in order to maintain leukemic gene expression in MLL-rearranged leukemia 2. Read the direction carefully: the fusion-driven, DOT1L-dependent state suppresses SIRT1's silencing activity. SIRT1 is a latent repressor that becomes active on leukemic loci only once DOT1L is pharmacologically removed.
Consequences of getting the direction right:
- SIRT1 belongs in this interactome as a downstream modifier of DOT1L-inhibitor response, not as a partner node. There is no interface node for it, and there should not be.
- No verified paper shows a direct SIRT1-MLL-fusion physical interaction. We searched for one and did not find it; nothing in this node asserts one, and no structure, residue range or affinity is offered for such a contact.
- On this model a SIRT1 inhibitor would be expected to make DOT1L inhibition worse, by removing the silencing the DOT1L drug unmasks. That is a prediction of one paper and the opposite has been observed in adjacent diseases - tenovin-6 kills primary paediatric ALL cells and sensitises them to etoposide and cytarabine, and SIRT1 knockout suppresses BCR-ABL leukaemogenesis [counter-evidence to E2]. See the counter-argument sweep section; the direction is disputed, not established.
The second SIRT1 paper in the verified set is narrower still: Sasca et al. report that SIRT1 prevents genotoxic-stress-induced p53 activation in acute myeloid leukemia, and the context of that work is FLT3-ITD, not KMT2A rearrangement 3. It is cited here at that scope and is not generalised.
The dependency number, and why its sign is the useful part
1:
| mean Chronos (1,208) | fraction depleted | leukemia (n=64) | KMT2A-r (n=10) | KMT2A::AFF1 (n=4) | selectivity vs leukemia | |
|---|---|---|---|---|---|---|
| SIRT1 | +0.035 | 0.010 | +0.077 | +0.116 | +0.097 | +0.039 |
+0.116 is the most positive KMT2A-rearranged mean of any gene in this dossier — more positive than MLLT3 (+0.059), CBX8 (+0.096) or BCOR (+0.003). Eight of ten KMT2A-rearranged lines score positive; only RS4;11 (-0.055) and THP-1 (-0.005) are negative, and both are within rounding of zero. Three of the four KMT2A::AFF1 lines are positive, with SEM at +0.224.
With n=10 the usual noise caveat applies and +0.039 selectivity relative to other leukemias carries no information. But the all-lines and KMT2A-rearranged means are both positive and consistent, and the sign is what matters here, because it agrees with the published mechanism rather than contradicting it. If SIRT1 is a silencer of the leukemic program, removing it should be neutral-to-mildly-favourable for a leukemic cell. That is exactly what the data show. This is the one gene in the batch where the DepMap sign and the literature mechanism corroborate each other.
Constraint is unremarkable by this dossier's standards: pLI 0.9779, LOEUF 0.497, 24 observed loss-of-function variants against 67.9 expected, mis_z 0.89 1. LOEUF 0.497 is the third-loosest in this dossier after EAF1 and EAF2 — SIRT1 is measurably less constrained than MLLT3 (0.172), BCOR (0.196) or KDM2B (0.240). HPA: low tissue specificity, detected in all tissues, group-enriched at protein level, nucleoplasmic.
Agonism or inhibition? Both compounds already exist, and the direction is the open question1 deeper
The genuinely actionable result in Chen et al. is the combination: SIRT1 activators combine with DOT1L inhibitors 2. That single sentence flips SIRT1 from a target to a lever, and it means the pharmacology this project should care about includes the ACTIVATOR column of ChEMBL, which is otherwise never read. Corrected 2026-09-19: it does not mean the inhibitor column can be ignored. The opposite direction has positive experimental results behind it in paediatric ALL and in BCR-ABL leukaemia [counter-evidence to E2], so the right experiment carries both arms and the heading of this section should be read as 'agonism, and also inhibition, head to head' rather than as a settled recommendation.
ChEMBL v34 target CHEMBL4506 ("NAD-dependent protein deacetylase sirtuin-1") has 215 activity rows at pChEMBL ≥ 6, best CHEMBL198609 at IC50 63 nM in a binding assay, and — unusually for this interactome — two curated mechanism drugs 4:
| Compound | ChEMBL | Mechanism | Max phase |
|---|---|---|---|
| SRT-2104 | CHEMBL4297436 | ACTIVATOR | phase 2 |
| EX-527 (selisistat) | CHEMBL382277 | INHIBITOR | phase 2 |
SRT-2104 is the compound the mechanism calls for, and it has already been through phase-2 clinical evaluation — meaning human safety and pharmacokinetic data exist. Combining an existing phase-2 activator with pinometostat-class DOT1L inhibition is about the cheapest clinically-informed combination experiment available anywhere in this dossier, and as far as we can verify nobody has reported it.
EX-527 was previously described here as the compound to avoid; that is withdrawn. It is the inhibitor-direction comparator the deciding experiment needs, and it is the one with the structural precedent: 4I5I at 2.50 Å is the SIRT1 catalytic domain 241-516 with NAD and an EX-527 analogue 5. Anyone who reaches for SIRT1 chemistry by picking the best-characterised co-crystal will pick the inhibitor, which by Chen et al.'s mechanism is the wrong direction - and by the paediatric-ALL and CML literature is the right one [counter-evidence to E2]. That is the whole dispute in one sentence.
One honest caution about the activator literature generally: 5BTR at 3.20 Å is SIRT1 with resveratrol and an AMC-containing peptide 5. The historical difficulty with sirtuin activators is that activation has often been substrate-dependent — apparent in assays using fluorophore-tagged peptides such as AMC-containing substrates and weaker or absent with native substrates. That the resveratrol complex was solved with an AMC peptide is a structural reminder of the problem, not a refutation of it, and it is the reason a SIRT1-activator combination experiment needs a native-substrate readout.
Structural coverage is good and mostly about regulation
Verified entries 5:
| PDB | Contents | Method / resolution | PMID |
|---|---|---|---|
4I5I | catalytic domain 241-516 + NAD + an EX-527 analogue | X-ray 2.50 Å | 23311358 |
4ZZJ | SIRT1 183-505 + activator + a p53 substrate peptide | X-ray 2.74 Å | 26134520 |
4ZZH | SIRT1 + activator | X-ray 3.10 Å | 26134520 |
5BTR | SIRT1 + resveratrol + an AMC peptide | X-ray 3.20 Å | 26109052 |
4IF6 / 4KXQ / 4IG9 | catalytic domain + C-terminal regulatory segment: 2.25 Å, 1.85 Å (closed), 2.64 Å (open) | X-ray | 24120939 |
The 4IF6/4KXQ/4IG9 set is the useful one for an activator programme, because it captures closed and open conformations of the catalytic domain with its C-terminal regulatory segment. An allosteric activator has to bias that equilibrium, and having both end states at 1.85-2.64 Å is a real asset. 4ZZJ, with an activator and a p53 substrate peptide bound simultaneously, is the single most informative entry for understanding how activation and substrate engagement are coupled.
That is a better structural position than most targets in this dossier enjoy — compare KDM2B, which has no structure of the domain its chemistry targets at all.
What this node does not claim
No direct SIRT1-KMT2A-fusion interaction. No verified source supports one; no structure, residue range, motif or affinity is offered for such a contact, and this node deliberately has no associated partner_interface node.
No Kd for any SIRT1-partner interaction. The only numeric potencies recorded are ChEMBL small-molecule activities 4.
The p53 result is kept at its published scope: SIRT1 prevents genotoxic-stress-induced p53 activation in AML, in an FLT3-ITD context 3. It is not extended to KMT2A-rearranged disease, and nothing here claims that a SIRT1 activator would be safe with respect to p53 signalling — if anything, the p53 result is a reason to worry about the activator direction, which is recorded in bite_back.
Mouse knockout phenotype is not established in this node; no verified citation was obtained this cycle.
Counter-argument sweep (2026-09-19): the direction is contested, not settled
This entry argues hard that the therapeutic direction for SIRT1 in KMT2A-rearranged leukaemia is agonism, on the strength of one paper (PMID 25822366) plus a DepMap sign. It had no counter-evidence recorded. There is a substantial literature on the other side, and in disease contexts closer to this project's than AML-plus-DOT1L-inhibitor.
In paediatric ALL, SIRT1 inhibition is the reported therapy. SIRT1 protein is higher in primary patient ALL cells than in healthy PBMC; the SIRT1/2 inhibitor tenovin-6 activated p53, killed pre-B ALL lines and primary cells from 41 children, reduced Mcl-1 and XIAP, blocked Wnt/beta-catenin signalling, eliminated CD133+CD19- ALL stem/progenitor cells and sensitised ALL cells to etoposide and cytarabine [counter-evidence to E2]. If that result generalises, a SIRT1 activator given alongside a genotoxic backbone would be the wrong drug in the wrong direction - which is the same risk this entry already flags from the p53 work in E3, now with a positive experimental result behind it rather than an inference.
In a fusion-driven leukaemia, SIRT1 has been reported as an oncogenic survival pathway. BCR-ABL upregulates SIRT1 through STAT5; SIRT1 knockout suppresses BCR-ABL transformation and CML-like disease, SIRT1 inhibition sensitises to imatinib, and tenovin-6 slows disease in mice [counter-evidence to E2]. A fusion oncoprotein driving SIRT1 expression is a different relationship from DOT1L restraining SIRT1 activity, and the two predict opposite signs.
What survives. Nothing found contradicts the specific epistatic claim in PMID 25822366 - that DOT1L inhibits SIRT1-mediated silencing, so SIRT1 activity is unmasked by a DOT1L inhibitor. That claim is narrow and may well be right. What does not survive is the framing of the direction as settled, and the implicit safety of the SRT-2104 combination.
A caveat that cuts against the counter-evidence too. Tenovin-6 inhibits SIRT1 and SIRT2, and a SirT2-preferring tenovin analogue raises p21 p53-independently (PMID 23322738), so tenovin phenotypes are not clean SIRT1 phenotypes. The honest summary is that this is a contested, context-dependent literature in which no compound used so far is selective enough to settle it - which is itself the most useful thing this entry can say.
Consequence for the recommended experiment. The SRT-2104 plus DOT1L-inhibitor combination should be run with a SIRT1-INHIBITOR arm alongside it, not instead of a vehicle arm. If SIRT1 inhibition also helps, the epistasis model is wrong and the entry's central recommendation should be withdrawn.
| uniprot | Q96EB6 |
|---|---|
| gene symbol | SIRT1 |
| length AA | 747 |
| protein name | NAD-dependent protein deacetylase sirtuin-1 |
| alphafold ID | AF-Q96EB6-F1 |
| alphafold verified | true |
| PDB verified against rcsb | true |
| subnuclear localization | NUCLEOPLASM plus an INDUCIBLE pool in PML NUCLEAR BODIES, recruited on PML or oncogenic Ras overexpression and colocalising there with p53 [E_q71_loc1]. The nucleoplasmic and PML-body annotations are separate UniProt entries at the same evidence level, so both pools are real. [EXPLICIT NEGATIVE] NOT MEASURED IN ANY KMT2A-REARRANGED CELL, and the PML-body result is an overexpression configuration in fibroblasts. UNPUBLISHED rather than not-applicable. |
domains
[
{
"name": "catalytic deacetylase domain (sirtuin)",
"start": 241,
"end": 516,
"note": "modelled as 241-516 in 4I5I with NAD and an EX-527 analogue; 4ZZJ models a wider 183-505"
},
{
"name": "C-terminal regulatory segment",
"start": null,
"end": null,
"note": "resolved together with the catalytic domain in 4IF6 (2.25 A), 4KXQ (1.85 A, closed) and 4IG9 (2.64 A, open); exact boundaries not supplied because no verified source gives them"
}
]PDB IDS
[ "4I5I", "4ZZJ", "4ZZH", "5BTR", "4IF6", "4KXQ", "4IG9" ]
key structures — the numbers behind the chart
[
{
"pdb": "4ZZJ",
"what": "SIRT1 183-505 + activator + p53 substrate peptide",
"method": "X-ray",
"resolution_A": 2.74,
"pmid": "26134520",
"title": "Crystallographic structure of a small molecule SIRT1 activator-enzyme complex.",
"authors": "Dai H, Case AW, Riera TV, et al.",
"venue": "Nat Commun",
"year": 2015,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
},
{
"pdb": "4KXQ",
"what": "SIRT1 catalytic + C-terminal regulatory segment, closed conformation",
"method": "X-ray",
"resolution_A": 1.85,
"pmid": "24120939",
"title": "Structural and functional analysis of human SIRT1.",
"authors": "Davenport AM, Huber FM, Hoelz A",
"venue": "J Mol Biol",
"year": 2014,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
},
{
"pdb": "4IG9",
"what": "SIRT1 catalytic + C-terminal regulatory segment, open conformation",
"method": "X-ray",
"resolution_A": 2.64,
"pmid": "24120939",
"title": "Structural and functional analysis of human SIRT1.",
"authors": "Davenport AM, Huber FM, Hoelz A",
"venue": "J Mol Biol",
"year": 2014,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
},
{
"pdb": "4I5I",
"what": "SIRT1 catalytic 241-516 + NAD + an EX-527 analogue (INHIBITOR — wrong direction for this biology)",
"method": "X-ray",
"resolution_A": 2.5,
"pmid": "23311358",
"title": "The 2.5 Å crystal structure of the SIRT1 catalytic domain bound to nicotinamide adenine dinucleotide (NAD+) and an indole (EX527 analogue) reveals a novel mechanism of histone deacetylase inhibition.",
"authors": "Zhao X, Allison D, Condon B, et al.",
"venue": "J Med Chem",
"year": 2013,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
},
{
"pdb": "5BTR",
"what": "SIRT1 + resveratrol + AMC-containing peptide",
"method": "X-ray",
"resolution_A": 3.2,
"pmid": "26109052",
"title": "Structural basis for allosteric, substrate-dependent stimulation of SIRT1 activity by resveratrol.",
"authors": "Cao D, Wang M, Qiu X, et al.",
"venue": "Genes Dev",
"year": 2015,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}
]essentiality
{
"depmap_common_essential": false,
"chronos_mean": 0.035,
"n_lines": 1208,
"fraction_lines_depleted": 0.01,
"chronos_mean_leukemia": 0.077,
"chronos_mean_kmt2ar": 0.116,
"chronos_mean_kmt2a_aff1": 0.097,
"selectivity_vs_leukemia": 0.039,
"selectivity_vs_all_lines": 0.081,
"gnomad_pLI": 0.9779,
"gnomad_loeuf": 0.497,
"gnomad_obs_lof": 24,
"gnomad_exp_lof": 67.9,
"gnomad_mis_z": 0.89,
"hpa_rna_tissue_specificity": "Low tissue specificity",
"hpa_rna_tissue_distribution": "Detected in all",
"hpa_rna_blood_cell_specificity": "Low immune cell specificity",
"mouse_ko": "not established in this node; no verified citation obtained this cycle",
"interpretation": "The most positive KMT2A-rearranged mean of any gene in this dossier: +0.116, with 8 of 10 lines positive. SIRT1 knockout is neutral-to-mildly-advantageous for KMT2A-rearranged cells. The +0.039 selectivity relative to other leukemias is noise at n=10; the load-bearing observation is the consistent positive SIGN, which corroborates PMID 25822366's mechanism that SIRT1 silences the leukemic program. Note the assay mismatch: an unperturbed CRISPR screen is the wrong experiment to detect a gene whose relevant activity is unmasked only by DOT1L inhibition, so this result neither supports nor refutes the DOT1L-inhibitor combination hypothesis."
}least depleted kmt2ar lines — the numbers behind the chart
[
{
"line": "SEM",
"fusion": "KMT2A::AFF1",
"chronos": 0.224
},
{
"line": "MOLM-14",
"fusion": "KMT2A::MLLT3",
"chronos": 0.223
},
{
"line": "HB11;19",
"fusion": "KMT2A::MLLT1",
"chronos": 0.214
}
]relationship to fusion
{
"type": "epistatic, downstream — NOT a physical partner",
"statement": "DOT1L inhibits SIRT1-mediated epigenetic silencing to maintain leukemic gene expression in MLL-rearranged leukemia (PMID 25822366). SIRT1 is a latent repressor that becomes active on leukemic loci once DOT1L is pharmacologically removed.",
"no_physical_interaction": "No verified paper shows a direct physical interaction between SIRT1 and any MLL/KMT2A fusion protein. No structure, residue range, motif or affinity is offered for such a contact, and this node deliberately has no associated partner_interface node.",
"therapeutic_direction": "CONTESTED as of 2026-09-19. PMID 25822366 predicts AGONISM: SIRT1 activators combine with DOT1L inhibitors because DOT1L restrains SIRT1-mediated silencing of the leukaemic programme. Against that, SIRT1 inhibition with tenovin-6 kills primary paediatric ALL cells and sensitises them to etoposide and cytarabine (PMID 25884180), and SIRT1 knockout suppresses BCR-ABL transformation and CML-like disease (PMID 22207735). No head-to-head activator-versus-inhibitor comparison exists in a KMT2A-rearranged model. The entry's earlier statement that a SIRT1 inhibitor 'would be expected to antagonise a DOT1L inhibitor' is a prediction of one model, not an observation, and the opposite has been observed in adjacent diseases."
}chemical matter
{
"chembl_target": "CHEMBL4506",
"chembl_pref_name": "NAD-dependent protein deacetylase sirtuin-1",
"n_activity_rows_pchembl_ge_6": 215,
"best_activity": {
"molecule": "CHEMBL198609",
"type": "IC50",
"value": 63,
"unit": "nM",
"assay_type": "B"
},
"curated_mechanism_drugs": [
{
"name": "SRT-2104",
"chembl": "CHEMBL4297436",
"mechanism": "ACTIVATOR",
"max_phase": 2,
"relevance": "the correct direction for this biology; human safety and pharmacokinetic data already exist"
},
{
"name": "EX-527 (selisistat)",
"chembl": "CHEMBL382277",
"mechanism": "INHIBITOR",
"max_phase": 2,
"relevance": "the WRONG direction; also the compound class in the best-characterised co-crystal, 4I5I"
}
],
"note": "The 215 potent rows and the 63 nM best value are inhibitor-direction potencies and are the wrong pharmacology for this mechanism. Counts are page-local (25-row API limit). max_phase 2 records that a compound reached phase 2, not that it succeeded. ChEMBL does not report whether SRT-2104's activation is substrate-dependent, which is the central historical criticism of sirtuin activators — 5BTR having been solved with an AMC-containing peptide is a structural reminder of it."
}interfaces
[]
direction dispute
{
"agonism_side": [
"25822366"
],
"inhibition_side": [
"25884180",
"22207735"
],
"selectivity_confound": "23322738 - tenovin-6 inhibits SIRT1 and SIRT2; a SirT2-preferring analogue raises p21 p53-independently, so tenovin phenotypes are not clean SIRT1 phenotypes",
"resolving_experiment": "SRT-2104 versus a selective SIRT1 inhibitor, head to head, with and without a DOT1L inhibitor, in KMT2A::AFF1 models, using a native-substrate target-engagement readout"
}depmap kmt2a selectivity
{
"chronos_mean_kmt2a_rearranged": 0.1101,
"chronos_mean_other_leukaemia": 0.0744,
"difference": 0.0357,
"pan_essentiality_all_lines": 0.0347,
"selectivity_rank": 13491,
"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."
}localization
{
"subcellular": "FIVE ANNOTATED LOCATIONS, ALL EXPERIMENTAL, INCLUDING A MITOCHONDRIAL ONE. UniProt Q96EB6 (fetched 2026-09-21) gives 'Nucleus, PML body', Cytoplasm, Nucleus, 'Nucleus, nucleoplasm' and Mitochondrion, every one at ECO:0000269. The mitochondrial pool matters for the chemistry: a NAD-dependent deacetylase with a mitochondrial pool sits in the same metabolic compartment as the sirtuins this entry warns it is confused with.",
"subnuclear": "NUCLEOPLASM plus an INDUCIBLE pool in PML NUCLEAR BODIES, recruited on PML or oncogenic Ras overexpression and colocalising there with p53 [E_q71_loc1]. The nucleoplasmic and PML-body annotations are separate UniProt entries at the same evidence level, so both pools are real.",
"explicit_no": "NOT MEASURED IN ANY KMT2A-REARRANGED CELL, and the PML-body result is an overexpression configuration in fibroblasts. UNPUBLISHED rather than not-applicable.",
"how_shown": "immunofluorescence colocalisation with PML and p53 under PML or Ha-rasV12 overexpression, plus deacetylation and senescence-rescue assays in primary MEFs (PMID 12006491); UniProt Q96EB6 SUBCELLULAR LOCATION comment, fetched 2026-09-21",
"systems": "mammalian cell lines and primary mouse embryo fibroblasts; no KMT2A-rearranged system",
"shown_for_fusion_localisation": "no",
"node": "localization-pml-nuclear-bodies-pmlrara"
}cell type
{
"evidence_class_key": "concept-expressed-required-maintains-three-evidence-classes",
"identifier_note": "SIRT1 (Q96EB6, 747 aa, entry name SIR1_HUMAN, synonym SIR2L1) is one of seven human sirtuins. SIRT2-SIRT7 are different proteins with different compartments - SIRT3, SIRT4 and SIRT5 are the mitochondrial sirtuins - and SIRT1 itself carries a mitochondrial annotation, which makes the family confusion easy to make and hard to notice. Almost all chemical matter in this area is reported at 'sirtuin' or 'SIRT1/2' level, so a compound result is usually a family result.",
"expressed_in": {
"class": 1,
"summary": "ubiquitous. SIRT1 is not a lineage-restricted protein."
},
"required_in": {
"class": 2,
"claim": "In MICE, SIRT1 is required for HOMEOSTATIC HAEMATOPOIETIC STEM CELL MAINTENANCE, and its deletion SKEWS DIFFERENTIATION TOWARD MYELOID AT THE EXPENSE OF THE LYMPHOID COMPARTMENT, with anaemia and an aging-like HSC phenotype, acting via FOXO3 [E_q71_ct1]. In a graph about a B-lineage disease, a gene whose loss collapses the lymphoid compartment is directly relevant - and it does so in an animal with no fusion.",
"pmids": [
"25068121"
],
"system_named_in_the_claim": "murine conditional SIRT1 deletion in normal haematopoiesis - no fusion present",
"nodes": [
"cell-type-hsc-bone-marrow",
"cell-type-clp"
]
},
"NOT_A_DEPENDENCY_CATEGORY_ERROR": {
"verdict": "SIRT1 IS NOT A DEPENDENCY IN EITHER ARM AND THE SCAN'S RANK DOES NOT APPLY TO IT. Chronos is POSITIVE IN BOTH ARMS: +0.1101 in KMT2A-rearranged and +0.0744 in other leukaemias, with pan-essentiality +0.0347 across all lines. Knocking SIRT1 out makes leukaemia lines grow slightly BETTER, everywhere.",
"why_this_matters": "This is the cleanest instance in this tranche of the category Q69 named for CDKN2A in cycle 39 - both arms positive - and it is cleaner than ATM, where only the index arm is positive. Rank 13,491 of 17,728 must NOT be counted as 'anti-selective', because ranking a non-dependency on a dependency scale is a category error. The entry's existing title, 'the least required gene in KMT2A-rearranged lines', is accurate as a description of the number and should be read as 'not required at all' rather than 'least required'.",
"and_it_bears_on_the_direction_dispute": "A gene whose knockout confers a small growth ADVANTAGE in every leukaemia arm is behaving as a brake in these lines. That is evidence on the side of the dispute this entry already records which says SIRT1 activation, not inhibition, is the therapeutic direction - though it is DepMap evidence, which is a proliferation readout in culture and not a disease claim."
},
"context": "murine normal haematopoiesis for the requirement; this project's DepMap scan for the (inapplicable) rank"
}cell state
{
"evidence_class_key": "concept-expressed-required-maintains-three-evidence-classes",
"maintains_state": {
"class": 3,
"states": [
"cell-state-stem-like-self-renewal",
"cell-state-quiescence"
],
"claim": "In MICE, SIRT1 maintains HSC HOMEOSTASIS via FOXO3, and its loss produces an AGEING-LIKE state - accumulated DNA damage, age-linked gene expression, myeloid-skewed output and anaemia [E_q71_ct1]. The state is a stem-cell homeostatic set point, which this graph represents through self-renewal and quiescence; neither node is an exact match and 'HSC ageing' has no node.",
"pmids": [
"25068121"
],
"system_named_in_the_claim": "murine normal haematopoiesis, conditional SIRT1 deletion"
},
"declined_vocabulary": "No 'HSC ageing' or 'myeloid-skewed output' state exists in this graph's 51-term cell vocabulary; recorded as a gap.",
"confidence": "strong",
"caveats": "Murine normal haematopoiesis. Nothing human, leukaemic, KMT2A-rearranged or infant. The two state nodes used are approximations of a homeostatic set point and are labelled as such. And SIRT1 is not a dependency in any leukaemia arm, so no state claim here supports a fusion-selective reading.",
"counter_evidence": {
"note": "The same protein rescues cells from PML/p53-induced premature senescence when overexpressed, which is the opposite functional direction from 'guardian of the stem-cell state'. The two results are not formally contradictory - one is loss-of-function in mouse HSCs, the other gain-of-function in fibroblasts - but they are why this entry's direction dispute exists, and neither can be quoted alone.",
"citation": {
"title": "Human SIR2 deacetylates p53 and antagonizes PML/p53-induced cellular senescence.",
"authors": "Langley E, Pearson M, Faretta M, Bauer UM, Frye RA, Minucci S, Pelicci PG, Kouzarides T",
"year": 2002,
"venue": "EMBO J",
"pmid": "12006491",
"doi": "10.1093/emboj/21.10.2383",
"type": "paper"
}
}
}depmap release backfill
{
"open_targets_dataversion": "26.06",
"note": "Added by Q71 per brief rule 83; Open Targets dataVersion fetched live 2026-09-21 as year 26 month 06. Scan denominator for this gene: n_all_lines 1208."
}- The direction is the trap. SIRT1 is habitually treated as an oncology target to inhibit, and the best-characterised co-crystal (4I5I) contains an inhibitor analogue. By PMID 25822366's mechanism a SIRT1 inhibitor should ANTAGONISE a DOT1L inhibitor. A programme that picks up SIRT1 without reading the direction will run the combination backwards and conclude the mechanism is wrong.
- The direction may simply be wrong, and the counter-literature is in the closer disease. Tenovin-6 (a SIRT1/2 inhibitor) killed primary ALL cells from 41 children, eliminated ALL stem/progenitor cells and sensitised to etoposide and cytarabine (PMID 25884180); SIRT1 knockout suppressed BCR-ABL leukaemogenesis and tenovin-6 slowed CML in mice (PMID 22207735). If that direction holds in KMT2A-rearranged disease, an SRT-2104 combination is not merely ineffective but actively protective of the leukaemia.
- Essential-function toxicity in the agonist direction: SIRT1 restrains genotoxic-stress-induced p53 activation in AML (PMID 24855208). Pharmacologically raising SIRT1 activity in a child receiving genotoxic chemotherapy could suppress a p53-dependent response and reduce the efficacy of the backbone regimen — a toxicity of efficacy rather than of organ damage, and one no single-agent assay detects.
- Compensation: SIRT2, SIRT3 and the other sirtuins share the NAD-dependent deacetylase mechanism, and the activity being manipulated depends on cellular NAD availability, which is itself metabolically regulated. An activator's effect may be capped by NAD supply rather than by target occupancy.
- Resistance: DepMap says SIRT1-null cells are not disadvantaged and KMT2A-rearranged lines gain slightly (+0.116). Under a DOT1L-inhibitor-plus-SIRT1-activator combination, loss of SIRT1 is a single-gene escape route that costs the leukemia nothing — and it is the one escape route the combination itself selects for.
- Delivery and assay validity: sirtuin activators have a long record of activity that depends on the assay substrate rather than on the enzyme. Without a native-substrate cellular readout and a target-engagement biomarker, a negative combination result would be uninterpretable and a positive one unconvincing.
- No compound in this literature is selective enough to settle the argument. Tenovin-6 hits SIRT1 and SIRT2, and a SirT2-preferring analogue raises p21 p53-independently (PMID 23322738), so both the pro-inhibition results and any refutation of them are confounded at the chemistry level before the biology is reached.
- SIRT1 deletion in mice collapses the LYMPHOID compartment while skewing output to myeloid (PMID 25068121). In a B-lineage disease treated in infants, a SIRT1 inhibitor is predicted to damage the lineage the patient needs to rebuild, and to do so through the normal HSC compartment.
- The dependency argument is not available in either direction. SIRT1's Chronos is POSITIVE in both DepMap arms (+0.1101 KMT2A-rearranged, +0.0744 other) with pan-essentiality +0.0347, so the gene is not required by any leukaemia line in the panel and its rank of 13,491 is a category error rather than a selectivity measurement.
- Sirtuin chemical matter is family-level. SIRT1 carries an experimental mitochondrial annotation alongside its nuclear ones and shares a catalytic mechanism with six paralogues, three of which are mitochondrial - so an activator or inhibitor result attributed to SIRT1 is usually a sirtuin-class result with false precision.
- Which direction is right in KMT2A-rearranged disease specifically? SIRT1 inhibition kills primary paediatric ALL cells and sensitises them to etoposide and cytarabine (PMID 25884180), and SIRT1 knockout suppresses BCR-ABL leukaemogenesis (PMID 22207735), while the DOT1L epistasis predicts the opposite (PMID 25822366). No experiment compares a SIRT1 activator and a SIRT1 inhibitor head to head in a KMT2A::AFF1 model, and until one does this entry's central recommendation rests on a single paper.
- Does SRT-2104, an existing phase-2 SIRT1 activator, combine with pinometostat-class DOT1L inhibition in KMT2A::AFF1 models? PMID 25822366 makes the prediction and both compound classes exist with human pharmacokinetic data. As far as we can verify the experiment has not been reported, and it is the cheapest clinically-informed combination in this dossier.
- Is SRT-2104's activation substrate-dependent? The historical criticism of sirtuin activators is that apparent activation depends on fluorophore-tagged peptide substrates and weakens or disappears with native substrates — and 5BTR was solved with an AMC-containing peptide. A combination experiment needs a native-substrate readout or its result will be uninterpretable.
- Does a SIRT1 activator blunt p53-dependent responses to conventional chemotherapy? SIRT1 prevents genotoxic-stress-induced p53 activation in AML (PMID 24855208, FLT3-ITD context). KMT2A-rearranged infant ALL is treated with genotoxic chemotherapy, so the combination that helps against a DOT1L inhibitor may hurt against cytarabine or an anthracycline. Nobody has measured the interaction.
- Which SIRT1 substrates carry the silencing effect at leukemic loci? PMID 25822366 establishes the epistasis but the node cannot name a substrate or a locus-level mechanism from verified sources, and without that there is no biomarker for whether an activator is engaging the right activity in a patient.
- Is there any direct physical connection between SIRT1 and a KMT2A fusion complex at all? We found no verified evidence for one. If the relationship is purely epistatic and runs through H3K79 methylation state, then SIRT1's recruitment to the relevant loci is an entirely separate and unaddressed question.
- Is any published SIRT1 phenotype in leukaemia attributable to SIRT1 alone? Tenovin-6 inhibits SIRT1 and SIRT2, and a SirT2-preferring analogue raises p21 p53-independently (PMID 23322738). A selective SIRT1 chemical probe, or an isoform-resolved genetic comparison, is the missing reagent for this whole question.
- Is SIRT1 in PML bodies at ENDOGENOUS expression, in any cell? Every image in the founding report follows overexpression of PML, Ras or SIRT1 itself, and this graph already records that punctate appearance under overexpression is weak evidence about the endogenous protein.
- Does the myeloid-versus-lymphoid skew on SIRT1 loss change the fate of a pre-leukaemic KMT2A-rearranged clone? The skew is measured in normal mice; the question of whether it shifts which lineage a fusion-bearing progenitor adopts has never been asked and is testable in the existing conditional model.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
SIRT1 is not DepMap common-essential; mean Chronos across 1,208 lines is +0.035 with 1.0% of lines depleted, the leukemia mean is +0.077, the 10 KMT2A-rearranged lines average +0.116 and the 4 KMT2A::AFF1 lines +0.097, giving selectivity_vs_leukemia +0.039 and selectivity_vs_all_lines +0.081; +0.116 is the most positive KMT2A-rearranged mean of any gene in this interactome dossier, with 8 of 10 lines positive; gnomAD gives pLI 0.9779, LOEUF 0.497, 24 observed versus 67.9 expected loss-of-function variants and mis_z 0.89; HPA classifies SIRT1 as low tissue specificity, detected in all tissues, low immune cell specificity, group-enriched at protein level, nucleoplasmic.
DepMap CRISPR (Chronos) gene effect via the Open Targets Platform GraphQL API; loss-of-function constraint via the gnomAD GraphQL API; normal-tissue expression via the Human Protein Atlas API · Broad Institute DepMap; gnomAD; Human Protein Atlas; Open Targets Platform · Databases 2026What this evidence cannot show
Means were computed by us from per-line API values rather than read from a DepMap release table, and the API does not state the release version. Only 10 KMT2A-rearranged lines (4 KMT2A::AFF1) contribute; the selectivity of +0.039 relative to other leukemias is deep inside the n=10 noise band of 0.15 and carries no information. What is load-bearing is the consistent positive sign across both the all-lines and the KMT2A-rearranged means, which agrees with the published mechanism that SIRT1 silences the leukemic program. Cell lines under standard culture are not cells under DOT1L inhibition, and the whole point of PMID 25822366 is that SIRT1's relevant activity is unmasked by a DOT1L inhibitor — so an unperturbed CRISPR screen is the wrong assay to detect it, and the flat-positive result neither supports nor refutes the combination hypothesis.
Evidence that argues against this (1)
- A mean of +0.035 with 1.0% of lines depleted, rising further in KMT2A-rearranged and KMT2A::AFF1 lines, is a negative result and the entry should lead with it. SIRT1 is not a dependency in this disease on this instrument. Oncofusion knowledge graph (internal analysis) 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DOT1L inhibits SIRT1-mediated epigenetic silencing in order to maintain leukemic gene expression in MLL-rearranged leukemia, so SIRT1 is a repressor acting downstream of DOT1L inhibition rather than a protein recruited by an MLL fusion; SIRT1 activators combine with DOT1L inhibitors.
DOT1L inhibits SIRT1-mediated epigenetic silencing to maintain leukemic gene expression in MLL-rearranged leukemia · Chen CW, Koche RP, Sinha AU, et al. · Nature Medicine 2015What this evidence cannot show
This establishes an epistatic relationship, not a physical one: NO verified paper shows a direct SIRT1-MLL-fusion physical interaction, and none is claimed here. The models are MLL-rearranged leukemia generally rather than KMT2A::AFF1 specifically. The combination claim is a preclinical result and the abstract reports no affinity or potency values, so the identity and pharmacology of the activators used cannot be transferred to SRT-2104 without checking the primary text. Directionally this result predicts that a SIRT1 inhibitor would antagonise a DOT1L inhibitor, which is the opposite of how SIRT1 is usually positioned in oncology.
Evidence that argues against this (2)
- The opposite therapeutic direction has been tested in the closest available disease context - paediatric ALL - and it worked. Jin et al. report that SIRT1 protein is HIGHER in primary ALL cells from patients than in peripheral blood mononuclear cells from healthy individuals, and that the SIRT1/2 INHIBITOR tenovin-6 activated p53, inhibited growth of pre-B ALL lines (REH, NALM-6) and of primary cells from 41 children and 2 adults with ALL, induced apoptosis, reduced Mcl-1 and XIAP, inhibited Wnt/beta-catenin signalling, eliminated CD133+CD19- ALL stem/progenitor cells, and SENSITISED ALL cells to etoposide and cytarabine. That is the exact opposite of this entry's 'agonism, not inhibition' thesis, in B-lineage ALL rather than in AML, and with primary paediatric material. Two honest qualifications: tenovin-6 inhibits SIRT1 and SIRT2 together, and a tenovin analogue selective for SirT2 raises p21 in a p53-INDEPENDENT manner (McCarthy et al., Mol Cancer Ther 2013, PMID 23322738), so tenovin phenotypes cannot be attributed to SIRT1 alone; and the ALL cohort is not selected for KMT2A rearrangement. But the direction question that this entry treats as settled is not settled, and if it is wrong the recommended SRT-2104 combination would be actively harmful. Jin Y et al. 2015
- In a fusion-oncoprotein leukaemia, SIRT1 has been reported as an oncogenic survival pathway rather than a latent silencer. Yuan et al. show BCR-ABL activating SIRT1 expression in haematopoietic progenitors via STAT5; SIRT1 activation promoted CML cell survival and proliferation with deacetylation of FOXO1, p53 and Ku70; SIRT1 inhibition sensitised CML cells to imatinib-induced apoptosis; SIRT1 knockout SUPPRESSED BCR-ABL transformation of mouse bone marrow and development of a CML-like myeloproliferative disease; and tenovin-6 deterred disease progression in mice. The mechanism is fusion-driven transcriptional upregulation of SIRT1, which is a different relationship from the DOT1L epistasis this entry records, and it predicts the opposite sign for the DepMap score in a fusion-driven leukaemia. The two results are not formally incompatible - different fusion, different lineage, and SIRT1's substrate repertoire is large - but they mean the literature on SIRT1 in leukaemia is genuinely contested and this entry should not present one direction as established. Yuan H et al. 2012
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
SIRT1 prevents genotoxic-stress-induced p53 activation in acute myeloid leukemia, in an FLT3-ITD context.
SIRT1 prevents genotoxic stress-induced p53 activation in acute myeloid leukemia · Sasca D, Hahnel PS, Szybinski J, et al. · Blood 2014What this evidence cannot show
Cited strictly at the paper's own scope: the context is FLT3-ITD AML, not KMT2A rearrangement, and it is not generalised here. It is included because it is the main reason to be cautious in the agonist direction — if SIRT1 activity restrains p53 activation under genotoxic stress, then pharmacologically increasing SIRT1 activity alongside conventional chemotherapy could blunt a p53-dependent response. That inference is ours and is recorded as a bite_back item and an open question rather than as a claim of this paper.
Evidence that argues against this (1)
- This is an FLT3-ITD AML context and the mechanism runs through p53. Infant KMT2A::AFF1 B-ALL is largely TP53 wild-type at diagnosis and rarely FLT3-ITD, so both conditions of the experiment are absent. Leukemia 2021
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
ChEMBL v34 target CHEMBL4506 (NAD-dependent protein deacetylase sirtuin-1) has 215 activity rows at pChEMBL >= 6 with a best value of IC50 63 nM (CHEMBL198609, binding assay), and two curated mechanism drugs: SRT-2104 (CHEMBL4297436), an ACTIVATOR at max phase 2, and EX-527 / selisistat (CHEMBL382277), an INHIBITOR at max phase 2.
ChEMBL database version 34, target CHEMBL4506 (NAD-dependent protein deacetylase sirtuin-1): target_search, get_bioactivity and get_mechanism queries · EMBL-EBI ChEMBL · ChEMBL 2026What this evidence cannot show
The 215 rows and the 63 nM best value are inhibitor-direction potencies and are the wrong pharmacology for the mechanism in PMID 25822366; the relevant record is the single ACTIVATOR. Counts and 'best' values are page-local under a 25-row API limit. A max_phase of 2 records that a compound reached phase 2, not that it succeeded; ChEMBL does not carry the outcome. ChEMBL does not report whether SRT-2104's activation is substrate-dependent, which is the central historical criticism of sirtuin activators and cannot be assessed from these rows.
Evidence that argues against this (1)
- A ChEMBL activity-row count measures how much medicinal chemistry has been DONE, not how tractable the target is, and the best reported affinity belongs to the most optimised compound in whichever series someone chose to pursue. Zero rows means nobody tried; thousands of rows means a target had a decade of industrial attention. SIRT1's 215 rows and two curated mechanism drugs come from a metabolic and ageing-biology programme, so the chemical matter exists for reasons unrelated to this disease and its selectivity was optimised against other endpoints. ChEMBL, EMBL-EBI 2026
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Verified SIRT1 structures include the catalytic domain 241-516 with NAD and an EX-527 analogue (4I5I, X-ray 2.50 A), SIRT1 183-505 with an activator and a p53 substrate peptide (4ZZJ, 2.74 A), SIRT1 with an activator (4ZZH, 3.10 A), SIRT1 with resveratrol and an AMC-containing peptide (5BTR, 3.20 A), and the catalytic domain with its C-terminal regulatory segment in both closed and open conformations (4IF6 2.25 A, 4KXQ 1.85 A closed, 4IG9 2.64 A open).
RCSB PDB entries 4I5I, 4ZZJ, 4ZZH, 5BTR, 4IF6, 4KXQ and 4IG9, with per-entity UniProt accessions and modelled residue ranges resolved against the RCSB search and data APIs · RCSB Protein Data Bank · RCSB PDB 2026What this evidence cannot show
Identifiers, resolutions and residue ranges were resolved against the RCSB API in this session; ranges are RCSB's UniProt-mapped modelled segments rather than curated domain boundaries. None of these entries involves a KMT2A fusion or any MLL-family protein, and none carries a published affinity for a protein-protein interaction. That 5BTR was solved with an AMC-containing peptide is a structural reminder of the substrate-dependent-activation criticism of sirtuin activators; the inference we draw from it is ours, not the paper's.
Evidence that argues against this (1)
- Counting PDB entries measures how tractable a protein was to CRYSTALLISE, which correlates with being small, ordered and well-behaved rather than with being druggable. What decides design value is whether the deposited density covers the site that matters, at a resolution that supports modelling, in the relevant state — and this graph has repeatedly found structures described in the direction that overstates what is resolved. RCSB PDB 2026
- E_q71_loc1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In MAMMALIAN CELLS, SIRT1 - the human Sir2 homolog - is RECRUITED TO PML NUCLEAR BODIES on overexpression of either PML or oncogenic Ha-rasV12, binds and deacetylates p53, co-localises with p53 in nuclear bodies when PML is upregulated, and when overexpressed in primary mouse embryo fibroblasts antagonises PML-induced p53 acetylation and RESCUES PML-mediated premature cellular senescence.
Human SIR2 deacetylates p53 and antagonizes PML/p53-induced cellular senescence. · Langley E, Pearson M, Faretta M, Bauer UM, Frye RA, Minucci S, Pelicci PG, Kouzarides T · EMBO J 2002What this evidence cannot show
Abstract read in full; full text not read. Mammalian cell lines and primary mouse embryo fibroblasts - not leukaemia, not haematopoietic, not KMT2A-rearranged (brief rule 73). CRITICAL LIMIT: the PML-body recruitment is observed UPON OVEREXPRESSION of PML or oncogenic Ras, and the rescue experiment overexpresses SIRT1 - so this describes an induced, overexpressed configuration, not the endogenous resting distribution. UniProt Q96EB6 independently annotates 'Nucleus, PML body', 'Nucleus, nucleoplasm', Nucleus, Cytoplasm and MITOCHONDRION, all at ECO:0000269 (fetched 2026-09-21).
Evidence that argues against this (1)
- The PML-body pool is one of at least five annotated compartments including a mitochondrial one, so treating 'SIRT1 is in PML bodies' as its address is a minority-pool error of the same shape this graph records for USP7, whose founding paper says only SOME of its foci colocalise with PML. In both cases the PML association is inducible and partial. Everett RD et al. 1997
- E_q71_ct1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In MICE, SIRT1 is required for HOMEOSTATIC HAEMATOPOIETIC STEM CELL MAINTENANCE and for LINEAGE SPECIFICATION: differentiation of young SIRT1-deleted HSCs is SKEWED TOWARD THE MYELOID LINEAGE with a significant DECLINE IN THE LYMPHOID COMPARTMENT and with anaemia; SIRT1-deleted HSCs accumulate damaged DNA and express age-linked molecules, giving a striking overlap with aged HSCs, and SIRT1 controls HSC homeostasis via the longevity transcription factor FOXO3.
Aging-like phenotype and defective lineage specification in SIRT1-deleted hematopoietic stem and progenitor cells. · Rimmelé P, Bigarella CL, Liang R, Izac B, Dieguez-Gonzalez R, Barbet G, Donovan M, Brugnara C, Blander JM, Sinclair DA, Ghaffari S · Stem Cell Reports 2014What this evidence cannot show
Abstract read in full; full text not read. MURINE conditional deletion in NORMAL haematopoiesis - not leukaemia, not human, not KMT2A-rearranged (brief rule 73). The 'aging-like' comparison is a transcriptional and phenotypic overlap, not a demonstration that SIRT1 loss causes ageing. The FOXO3 mechanism is stated at abstract level without an epistasis experiment being described there.
Evidence that argues against this (1)
- The direction of SIRT1 in disease is genuinely disputed and this entry already records that dispute. The same protein that is required for normal HSC homeostasis also antagonises p53 and RESCUES cells from PML/p53-induced senescence when overexpressed - which is an oncogenic direction. A deacetylase that is both needed for normal stem-cell homeostasis and able to suppress a senescence barrier cannot be assigned a single therapeutic direction from these results, and this project's own scan shows SIRT1 is not a dependency in any leukaemia arm. Langley E et al. 2002
- regulates →KMT2A / MLL1 — the gene that breaks
opposes KMT2A function rather than partnering with it
recorded as the least-required gene in the KMT2A-r panel; a negative reference point for dose claims
named as the deacetylase for this mark in the MLL-AF9 SIRT1/DOT1L study
deacetylates H3K9 at MLL-AF9 target TSSs after DOT1L inhibition; Sirt1 knockdown blocks the loss
Counter-evidence (Jin 2015, PMID 25884180): SIRT1 is higher in primary ALL cells than in healthy PBMC, and the SIRT1/2 INHIBITOR tenovin-6 activates p53, kills primary cells from 41 children with ALL, eliminates CD133+CD19- ALL stem/progenitor cells and sensitises to etoposide and cytarabine - the opposite direction from this entry's agonism thesis.
Direction dispute recorded 2026-09-19: the DOT1L-epistasis model (PMID 25822366) calls for a SIRT1 ACTIVATOR; SIRT1 inhibition kills primary paediatric ALL cells (PMID 25884180) and SIRT1 knockout suppresses BCR-ABL leukaemogenesis (PMID 22207735). No head-to-head comparison exists in a KMT2A-rearranged model, and tenovin-6 is not SIRT1-selective (PMID 23322738).
same paper; SIRT1 is graded bait-supported and CLOCK prey-only from the same figure set
SIRT1 is the specific eraser; SIRT2/3/6/7 and class I/II HDACs are negatives
SIRT1 deacetylates K1130; SIRT2/3/6/7 and class I/II HDACs do not
SIRT1 deacetylates K1133; catalytically dead SIRT1(H363Y) does not
inside the mapped SIRT1-contact region 650-1327 but never tested as a substrate
the one unread hit is a SIRT1-in-leukaemia-stem-cells review
EX527 arm as the movability control
SIRT1 is recruited to PML nuclear bodies on PML or oncogenic Ras overexpression and colocalises there with p53, antagonising PML-induced p53 acetylation - an INDUCED, overexpressed configuration, and one of five experimentally annotated compartments (mammalian cells and primary MEFs; PMID 12006491)
REQUIRED IN: SIRT1 is required for homeostatic haematopoietic stem cell maintenance via FOXO3, and its deletion gives an aging-like HSC phenotype - murine normal haematopoiesis, no fusion present (PMID 25068121)
SIRT1-deleted HSCs skew myeloid with a significant DECLINE IN THE LYMPHOID COMPARTMENT - directly relevant to a B-lineage disease, and measured in an animal with no fusion (murine; PMID 25068121)
- regulates →The stem-like self-renewal state — a limited stem-cell module grafted onto a committed cell
SIRT1 maintains the HSC homeostatic set point via FOXO3; its loss produces accumulated DNA damage and age-linked expression (murine normal haematopoiesis; PMID 25068121)
same minority-pool trap: both proteins are best known subnuclearly for a PARTIAL, inducible association with PML nuclear bodies while carrying four to five other experimental compartments, and in both cases the argued pool is not the imaged one (PMIDs 12006491, 9034339)
third menin-lipid mechanism in the same tissue (PMID 23867312)
SIRT1 inhibition prevents the H4K16ac loss that follows Mll1 deletion
- ← related toLost from the KMT2A half: the MOF/KAT8 contact - required in the cell, but not shown to matter to the fusion
SIRT1 opposes the MLL1-MOF axis; its inhibition restores H4K16ac at KMT2A targets