FACT (SPT16-SSRP1) - the chromatin arm of transcription elongation, and a clean negative result for the KMT2A::AFF1 interactome
also called FACT · FACT complex · facilitates chromatin transcription · SPT16-SSRP1 · SUPT16H-SSRP1 heterodimer
FACT is a two-protein machine that pulls nucleosomes apart in front of RNA polymerase and puts them back behind it, so genes packed into chromatin can be read. This entry exists to record, with numbers, that despite being central to transcription it has no documented role in the KMT2A fusion machinery - which is a useful thing for a knowledge graph to state plainly rather than leave as an unexplored gap.
The same thing, for a scientist
FACT is the obligate SPT16-SSRP1 heterodimer that permits Pol II transcription through chromatin, binds nucleosomes and H2A-H2B dimers, and loses activity when nucleosomal histones are crosslinked [E1]. This graph's elongation axis is otherwise densely covered - SEC, P-TEFb, DSIF, NELF, LEC, ELL-EAF, PAF1C - and FACT was the conspicuous absence. **The absence is in the field, not only in the graph.** Across a deliberately narrow, title-checked query set: 1 record for SPT16 against MLL/KMT2A, 1 for SSRP1, 1 relevant record for the complex, **0 for FACT against AFF1/AFF4/ELL2/SEC/MLLT1/MLLT3 after title-check**, and **0 for FACT against DOT1L or H3K79** [E2]. Dependency says the same thing from the other direction: both subunits are pan-essential (SPT16 -1.763, SSRP1 -1.2714) and both are **anti-dependencies** in KMT2A-rearranged lines (+0.1165 rank 17,162 and +0.2651 rank 17,672 of 17,728), with no requirement in either lineage half of the index panel [E3][E4]. Two threads remain and both are recorded as threads: **ENL/MLLT1 interacts with FACT and is required for FACT deposition at cis-regulatory elements** (colorectal cancer, wild-type ENL) [E5], and **LEDGF/p75 binds SSRP1 through its PWWP domain**, LEDGF being an established component of the menin-MLL tethering module [E6]. The only FACT-directed agent with MLL-r data, the curaxin CBL0137, works by **trapping** FACT in chromatin and acts through p53 and interferon [E7] - a gain-of-occupancy mechanism that argues nothing about whether FACT function is required.
What this complex entry adds beyond the two subunit entries
Three things that can only be said at complex level.
1. The subunits disagree on a statistic they cannot biologically disagree on. SPT16 and SSRP1 are an obligate heterodimer, yet their KMT2A-selectivity differences are +0.1165 and +0.2651 - more than twofold apart, and 510 rank positions apart 3. Since neither protein can have a requirement independent of the other, the spread is a direct measurement of how much noise this statistic carries at the essentiality floor. That is a calibration fact the graph can reuse: for any pan-essential obligate complex, treat a difference of ~0.15 as indistinguishable from zero.
2. The two structural holes this beat was sent to close are not the same kind of hole. The brief grouped FACT and NSL together as 'structural holes'. NSL turned out to have direct, mapped, published contacts with MLL1 and with NUP98 fusions. FACT turned out to have none. Writing both as entries makes the difference visible; writing neither, or writing a single 'chromatin cofactors' entry, would have hidden it.
3. The question the dependency data cannot answer is still open. A CRISPR viability screen asks 'does the cell die'. The interesting FACT question is 'is nucleosome turnover at KMT2A::AFF1 target genes different from turnover at matched non-target genes', which is a nascent-transcription and occupancy question. Nothing in this entry closes it, and the open questions say so.
The two live threads, and why neither is yet a finding
Thread 1 - ENL. ENL (MLLT1) interacts with FACT and is required for FACT deposition over cis-regulatory elements 5. ENL is a KMT2A fusion partner and a SEC subunit, so this is the only published route by which FACT could touch the fusion axis. The work is in colorectal cancer with wild-type, overexpressed ENL; no fusion is tested; and the SEC query still returns zero 2. The testable version is narrow and cheap: co-immunoprecipitate FACT from a KMT2A::MLLT1 line, which retains the ENL YEATS domain.
Thread 2 - LEDGF/p75. LEDGF complexes with FACT specifically through SSRP1, hSpt16-independently, using the LEDGF PWWP domain and the SSRP1 HMG domain 6. LEDGF's other relevant module, the IBD, is the one that engages the menin-MLL complex. Because the contacts use different LEDGF domains, a ternary arrangement is not excluded by geometry. It is, however, entirely untested, and LEDGF is a promiscuous adaptor whose interactome is known to switch with phosphorylation state.
Both threads share a shape the graph should be wary of: a real, mapped contact in a non-leukaemia system, connected to the fusion axis by one extrapolation. That shape has produced good leads and bad entries in roughly equal numbers in this project.
Cell type, cell state, localisation, epigenetics - stated at complex level
Subcellular: nucleus and chromosome for both subunits. Subnuclear: SSRP1 is additionally annotated nucleolar and SPT16 is not; within an obligate heterodimer that is an unresolved asymmetry, and this entry records it as unresolved. No subnuclear localisation of FACT has been reported in any KMT2A-rearranged system.
Cell type: in this disease, 12 MLL-r leukaemia cell lines and MLL-r ALL patient-derived xenograft cells - and in every case as the target of a drug, never as a protein under genetic study 7. Cell state: not established. There is no differentiation, quiescence or leukaemia-stem-cell readout for FACT in KMT2A-rearranged disease.
Epigenetics: FACT writes no mark and has no catalytic activity. Its epigenetic quantity is nucleosome occupancy and H2A-H2B dimer exchange around elongating Pol II 1. Note the direction of the only available pharmacology: CBL0137 increases FACT's chromatin occupancy 7. An entry that listed 'FACT inhibitor' under this complex would be describing the opposite of the published tool compound.
Condensates: what this entry deliberately does not say
The FACT literature contains phase-separation claims, and the ENL work this entry cites sits next to a substantial ENL-condensate literature that this graph already carries. Nothing in the evidence read for this entry used calibrated half-bleach FRAP measuring internal mixing, which is the only assay that distinguishes liquid-liquid phase separation from binding to spatially clustered sites; 1,6-hexanediol would not settle it either, and in chromatin it rapidly immobilises and condenses chromatin rather than melting droplets.
So this entry makes no phase-separation claim about FACT, in either direction. 'FACT is deposited over cis-regulatory elements' 5 is a chromatin-occupancy statement and is compatible with clustered-site binding, which is the rival model and, at regulatory elements, arguably the more interesting one because it predicts sequence-specific targeting.
| complex name | FACT (FAcilitates Chromatin Transcription) |
|---|---|
| stoichiometry | obligate heterodimer, SPT16 + SSRP1 |
| classification | general transcription-through-chromatin machinery; NOT a documented component of the KMT2A::AFF1 interactome |
| position in mechanism | nucleosome chaperone travelling with elongating Pol II; sits alongside but is not documented to contact the SEC, DSIF, NELF, LEC or PAF1C modules this graph already carries |
| cell type evidence | 12 MLL-r leukaemia cell lines and MLL-r ALL PDX cells, pharmacological only |
| cell state evidence | not established |
| epigenetic readout | no mark written; nucleosome occupancy and H2A-H2B dimer exchange. The only tool compound INCREASES FACT chromatin occupancy rather than reducing it. |
| condensate stance | no phase-separation claim is made in either direction; no calibrated half-bleach internal-mixing measurement exists in the evidence read, and hexanediol/classical FRAP could not settle it |
| granularity exception | This is a complex-level entry that exists alongside, not instead of, separate entries for both subunits (gene-supt16h, gene-ssrp1). It carries only the three facts that cannot be stated at subunit level: the within-heterodimer disagreement on the selectivity statistic, the contrast with NSL as the other 'structural hole' in this beat, and the complex-level condensate stance. |
subunits
[
{
"gene_symbol": "SUPT16H",
"uniprot": "Q9Y5B9",
"length_aa": 1047,
"aliases": [
"SPT16",
"FACT140",
"FACTP140",
"CDC68"
],
"separate_entry": true,
"node": "gene-supt16h",
"depmap": {
"difference": 0.1165,
"rank": 17162,
"percentile": 96.807,
"mean_kmt2a_r": -1.8124,
"mean_other": -1.9288,
"pan": -1.763
},
"lineage_split": {
"raw_mean_kmt2a_myeloid": -1.958,
"z_kmt2a_myeloid": -0.11,
"raw_mean_kmt2a_b_lineage": -1.639,
"z_kmt2a_b_lineage": 0.86
}
},
{
"gene_symbol": "SSRP1",
"uniprot": "Q08945",
"length_aa": 709,
"aliases": [
"FACT80",
"T160"
],
"separate_entry": true,
"node": "gene-ssrp1",
"depmap": {
"difference": 0.2651,
"rank": 17672,
"percentile": 99.684,
"mean_kmt2a_r": -1.1468,
"mean_other": -1.412,
"pan": -1.2714
},
"lineage_split": {
"raw_mean_kmt2a_myeloid": -1.233,
"z_kmt2a_myeloid": 0.54,
"raw_mean_kmt2a_b_lineage": -1.118,
"z_kmt2a_b_lineage": 0.9
}
}
]localization
{
"subcellular": "nucleus; chromosome",
"subnuclear": "SSRP1 annotated nucleolar, SPT16 not - unresolved asymmetry within an obligate heterodimer; nothing reported in any KMT2A system"
}drugs
[
{
"name": "CBL0137 (curaxin)",
"mechanism": "traps FACT into chromatin; p53 activation and interferon response",
"mll_r_data": "12 lines, 3 PDX populations, MV4;11 subcutaneous model, 5 MLL-r ALL PDX models",
"evidence": "E7",
"caution": "gain-of-occupancy tool; not evidence of a FACT loss-of-function dependency"
}
]live threads
[
{
"thread": "ENL/MLLT1 interacts with FACT and is required for FACT deposition at cis-regulatory elements",
"system": "colorectal cancer, wild-type ENL",
"evidence": "E5",
"test": "co-IP FACT from a KMT2A::MLLT1 line, which retains the ENL YEATS domain"
},
{
"thread": "LEDGF/p75 PWWP binds the SSRP1 HMG box, hSpt16-independently; LEDGF IBD separately engages menin-MLL",
"system": "HIV-1 / CD4+ T cells",
"evidence": "E6",
"test": "sequential IP from a KMT2A::AFF1 line asking whether LEDGF can hold SSRP1 and menin simultaneously"
}
]enumeration record counts — the numbers behind the chart
[
{
"query": "(MLL OR KMT2A) AND (SUPT16H OR SPT16 OR FACP140 OR FACTP140)",
"records": 1
},
{
"query": "(MLL OR KMT2A) AND (SSRP1 OR T160 OR FACT80)",
"records": 1
},
{
"query": "(MLL OR KMT2A) AND (SUPT16H OR SPT16 OR FACTP140 OR \"FACT complex\" OR SSRP1 OR \"T160\" OR FACT140)",
"records": 2,
"records_after_title_check": 1
},
{
"query": "(SUPT16H OR SSRP1 OR \"FACT complex\") AND (AFF4 OR AFF1 OR ELL2 OR \"super elongation complex\" OR MLLT1 OR MLLT3)",
"records": 1,
"records_after_title_check": 0
},
{
"query": "(SUPT16H OR SSRP1 OR \"FACT complex\") AND (DOT1L OR H3K79)",
"records": 0
},
{
"query": "\"FACT complex\" AND (leukemia OR leukaemia)",
"records": 9,
"records_after_title_check": 0
},
{
"query": "(SUPT16H OR SPT16) AND (leukemia OR leukaemia)",
"records": 5,
"records_after_title_check": 1
},
{
"date": "2026-09-19"
}
]- Both subunits are pan-essential across 1,208 lines with no KMT2A window. Any FACT-directed intervention is a general transcription toxin unless it exploits something other than loss of function.
- The one clinical-stage agent, CBL0137, works through p53 - so it should fail precisely in TP53-altered disease, which is the highest-risk subset, and the published MLL-r work does not test that.
- Because SPT16 and SSRP1 are obligate partners there is no subunit-selective handle, and no way to spare one tissue by choosing a subunit.
- A degrader strategy would be reasoning backwards from a trapping agent. Removing FACT and trapping FACT are opposite perturbations and the genetic data for removal is discouraging.
- Is nucleosome turnover at KMT2A::AFF1 target genes different from turnover at expression-matched non-target genes? This is the question a viability screen cannot ask and the one that would actually decide whether FACT is mechanistically relevant here.
- Does a KMT2A::MLLT1 fusion recruit FACT through its retained ENL YEATS domain? One co-immunoprecipitation would convert the only live thread into a result or kill it.
- Can LEDGF/p75 engage SSRP1 and the menin-MLL module at the same time? The domains differ (PWWP versus IBD), so the answer is not forced by geometry and nobody has looked.
- Does the twofold disagreement between two obligate subunits on the selectivity statistic generalise? If it does, it sets a floor of roughly 0.15 below which no difference in this scan should be interpreted, which would retrospectively affect several entries in this graph.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
FACT, the chromatin-specific elongation activity, comprises the human Spt16/Cdc68 homologue and SSRP1; it specifically interacts with nucleosomes and H2A/H2B dimers and its activity is abrogated by covalently crosslinking nucleosomal histones.
The chromatin-specific transcription elongation factor FACT comprises human SPT16 and SSRP1 proteins · Orphanides G, Wu WH, Lane WS, Hampsey M, Reinberg D · Nature 400(6741):284-8 1999What this evidence cannot show
Abstract fetched from PubMed and read on 2026-09-19; full text not read and every identifier copied from the fetched record. In vitro reconstitution only; no cellular, haematopoietic or KMT2A system. The 1999 nucleosome-disassembly interpretation has been revised by later structural work and should not be quoted as the current model.
Evidence that argues against this (1)
- Whether FACT's dominant in vivo role is disassembly ahead of polymerase or reassembly behind it remains contested, and the two models predict opposite consequences for a gene whose nucleosomes turn over rapidly. Nothing in this entry depends on the answer, but a reader importing 'FACT removes nucleosomes' into a mechanism diagram would be over-reading a 1999 crosslinking result. Li Y et al. 2023
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
As of 2026-09-19 PubMed returns, for deliberately narrow and individually title-checked queries: 1 record for '(MLL OR KMT2A) AND (SUPT16H OR SPT16 OR FACP140 OR FACTP140)'; 1 for '(MLL OR KMT2A) AND (SSRP1 OR T160 OR FACT80)'; 2 for the combined FACT query of which 1 survives title-check; 1 for '(SUPT16H OR SSRP1 OR "FACT complex") AND (AFF4 OR AFF1 OR ELL2 OR "super elongation complex" OR MLLT1 OR MLLT3)' which is a 2002 PITSLRE kinase paper and does not survive title-check, giving 0; 0 for '(SUPT16H OR SSRP1 OR "FACT complex") AND (DOT1L OR H3K79)'; and 9 for '"FACT complex" AND (leukemia OR leukaemia)' of which 0 concern KMT2A on title-check.
PubMed E-utilities esearch record counts for the FACT complex and its subunits against KMT2A, MLL, the super elongation complex and DOT1L, run for interactome round 5 · NCBI PubMed, queried 2026-09-19 2026What this evidence cannot show
Counts from the esearch `count` field; **every record in every counted query above was title-checked individually**, which is why several corrected counts are lower than the raw counts. Symbol set: SUPT16H, SPT16, FACP140, FACTP140, FACT140, SSRP1, T160, FACT80, plus the phrase 'FACT complex'. Not searched as phrases: 'facilitates chromatin transcription', 'Cdc68', 'structure-specific recognition protein'. A counted absence is a floor on recall and the deliberate narrowness makes these numbers conservative in that direction.
Evidence that argues against this (1)
- Both subunit names are hostile to text search in opposite ways: 'FACT' is an ordinary English word (inflating), while 'facilitates chromatin transcription' spelled out with no gene symbol would be missed entirely (deflating). This is the precision-and-recall pair named in the standing alias rule arriving in a single gene, and it means these counts bound the literature loosely rather than tightly. Broad Institute DepMap / Open Targets Platform; analyses/depmap-kmt2a-selective-scan-full 2026
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
In the genome-wide KMT2A-rearranged selectivity scan of 17,728 genes, both FACT subunits are pan-essential and anti-selective: SUPT16H pan -1.763, difference +0.1165, rank 17,162 (96.807th percentile), KMT2A-rearranged mean -1.8124 against other-leukaemia -1.9288; SSRP1 pan -1.2714, difference +0.2651, rank 17,672 (99.684th percentile), KMT2A-rearranged mean -1.1468 against other-leukaemia -1.412.
Genome-wide KMT2A-rearranged selectivity scan, 17,728 genes (DepMap Chronos via Open Targets) · Broad Institute DepMap / Open Targets Platform; analyses/depmap-kmt2a-selective-scan-full 2026What this evidence cannot show
Rows read out of analyses/depmap-kmt2a-selective-scan-full/results.json on 2026-09-19 in a scripted pass. Eleven KMT2A-rearranged against 43 other leukaemia lines, AML-heavy, none infant-derived. **Both genes sit near the Chronos floor where the difference statistic is compressed and its sign unreliable**; the correct reading is 'no selective requirement', not 'protective'.
Evidence that argues against this (1)
- The two subunits of an obligate heterodimer differ by 0.1486 on this statistic and by 510 rank positions, which is a direct internal measurement of the statistic's noise at the essentiality floor and cautions against reading any difference of similar size elsewhere in this beat as signal. Broad Institute DepMap / Open Targets Platform; analyses/depmap-interactome5-lineage-split 2026
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
On the lineage-split run for this beat, neither FACT subunit shows a KMT2A-rearranged requirement in either lineage half. SUPT16H: myeloid raw mean -1.958 (z -0.11), B-lineage raw -1.639 (z +0.86), other leukaemia -1.921 (SD 0.327). SSRP1: myeloid raw -1.233 (z +0.54), B-lineage raw -1.118 (z +0.90), other leukaemia -1.404 (SD 0.317). Both are classified 'weak or ambiguous' and neither is flagged as a lineage artefact.
Interactome round 5 panel, lineage-split DepMap dependency (FACT, NSL, MSL, COMPASS remainder, Mediator) · Broad Institute DepMap / Open Targets Platform; analyses/depmap-interactome5-lineage-split 2026What this evidence cannot show
Rows read from analyses/depmap-interactome5-lineage-split/results.json on 2026-09-19. Six myeloid and four B-lineage index lines per gene, so a single line moves a mean; split is by lineage LABEL, not expression signature. Raw means are given beside every z because the comparator SDs here (0.317-0.327) are small precisely because all leukaemia lines need FACT, which inflates z relative to the underlying difference.
Evidence that argues against this (1)
- With six and four index lines this analysis can exclude only a large KMT2A-specific requirement. A modest one - the size that would matter mechanistically without showing up in a viability screen - remains entirely possible and would need a different experiment to exclude. Broad Institute DepMap / Open Targets Platform; analyses/depmap-kmt2a-selective-scan-full 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
ENL, encoded by MLLT1, interacts with the histone chaperone FACT complex and is required for the deposition of FACT over cis-regulatory elements, a process regulated by BRD4; ENL occupies the majority of super-enhancers across cancer types and its auxin-degron depletion decommissions them and represses super-enhancer-controlled oncogenes including MYC.
Targeting the Epigenetic Reader ENL Inhibits Super-Enhancer-Driven Oncogenic Transcription and Synergizes with BET Inhibition to Suppress Tumor Progression · Chen Y, Ying Y, Ma W, Ma H, Shi L, Gao X, Jia M, Li M, Song X, Kong W, Chen W, Zheng X, Muluh TA, Wang X, Wang M, Shu XS · Cancer Research 84(8):1237-1251 2024What this evidence cannot show
Abstract fetched from PubMed and read on 2026-09-19; full text not read, identifiers copied from the fetched record. **Colorectal cancer, wild-type overexpressed ENL, no KMT2A fusion anywhere in the work.** The interaction is asserted at abstract level without the assay being specified in what was read.
Evidence that argues against this (1)
- Directly opposed by the counted absence: zero records connect FACT to AFF1, AFF4, ELL2 or the SEC after title-check, so this contact is currently a property of ENL outside the fusion context and not of the fusion machinery. NCBI PubMed, queried 2026-09-19 2026
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
LEDGF/p75 complexes with FACT; the interaction is with SSRP1 in an hSpt16-independent manner and requires the PWWP domain of LEDGF proteins and the HMG domain of SSRP1.
The Structure-Specific Recognition Protein 1 Associates with Lens Epithelium-Derived Growth Factor Proteins and Modulates HIV-1 Replication · Lopez AP, Kugelman JR, Garcia-Rivera J, Urias E, Salinas SA, Fernandez-Zapico ME, Llano M · Journal of Molecular Biology 428(14):2814-31 2016What this evidence cannot show
Abstract fetched from PubMed and read on 2026-09-19; full text not read and identifiers copied from the fetched record. HIV-1 biology in CD4+ T cells; no leukaemia, no KMT2A, no fusion. The claim that a ternary LEDGF-FACT-menin-MLL arrangement is geometrically permitted follows from the domains used and is an inference of this entry, not a finding of the paper.
Evidence that argues against this (1)
- LEDGF/p75 is a promiscuous chromatin adaptor whose IBD interactome is documented to switch with kinase-dependent phosphorylation, so a mapped contact in one cell state is weak evidence about another. The graph should hold this as a testable geometry, not as a pathway. PubMed record; title and year read from the esummary response for this PMID on 2026-09-19. Authors, journal and DOI were NOT fetched in this session and are deliberately omitted rather than written from recall. 2018
- E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The curaxin CBL0137 exerts its anticancer effect by trapping FACT into chromatin, activating the p53 pathway and inducing an interferon response; it reduced viability of 12 MLL-rearranged leukaemia cell lines and 3 MLL-r ALL patient-derived xenograft cell populations with submicromolar IC50s and reduced leukaemia burden in a subcutaneous MV4;11 model and 5 MLL-r ALL xenograft models.
Potent antileukemic activity of curaxin CBL0137 against MLL-rearranged leukemia · Somers K, Kosciolek A, Bongers A, El-Ayoubi A, Karsa M, Mayoh C, Wadham C, Middlemiss S, Neznanov N, Kees UR, Lock RB, Gudkov A, Sutton R, Gurova K, Haber M, Norris MD, Henderson MJ · International Journal of Cancer 146(7):1902-1916 2020What this evidence cannot show
Abstract fetched from PubMed and read on 2026-09-19; full text not read, identifiers copied from the fetched record. **Trapping is gain of occupancy, not inhibition**, so this result does not establish that FACT function is required in MLL-r cells - the genetic data indicates it is not selectively required. No non-MLL-r leukaemia comparator arm appears in the abstract, so MLL-r selectivity is not established either.
Evidence that argues against this (1)
- Curaxin activity is reported in leukaemias driven by wholly different lesions, including JAK2 V617F disease and BCR::ABL1 chronic myeloid leukaemia, which is exactly what a p53/interferon mechanism predicts and which undercuts any KMT2A-specific reading of the MLL-r efficacy data. Pearson S et al. 2022
two elongation-coupled histone chaperones travelling with Pol II
FACT subunits are the LEAST required genes in the KMT2A-rearranged arm of this scan
Large subunit of the FACT histone chaperone.
Small subunit of the FACT histone chaperone.
SPT16 is the large subunit.
SSRP1 is the small subunit.
ENL interacts with FACT and is required for its deposition at cis-regulatory elements - colorectal cancer, wild-type ENL.
LEDGF/p75 PWWP binds the SSRP1 HMG box; LEDGF's IBD separately engages the menin-MLL module.
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
Clean negative: zero title-checked records link FACT to the fusion or the SEC, and both subunits are anti-dependencies.
Sits on the same elongation axis the graph covers through LEC, DSIF, NELF and SEC, and is the chromatin-chaperone arm none of those entries covers.
The counted-absence half of the beat: zero title-checked records link FACT to the fusion axis.
the two FACT subunits disagree at p=0.01 and p=0.29
FACT is enriched at MLL-AF4 target enhancers