Oncofusion

start / graph / gene-hexim1

Proteinstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q35-ptefb-reservoir

HEXIM1 - the brake pad, the only part of the reservoir anyone has deliberately raised in a KMT2A-rearranged cell, and the result was that the leukaemia grew less

also called HEXIM1 · CLP1 · CLP-1 · EDG1 · HIS1 · MAQ1 · hexamethylene bisacetamide-inducible protein 1

HEXIM1 is the protein that actually switches P-TEFb off inside the 7SK particle - the RNA only holds the two together. When researchers forced KMT2A-rearranged AML cells to make more HEXIM1, the cells grew more slowly, and when they removed it the cells grew faster and made more MYC.

The same thing, for a scientist

HEXIM1 (UniProt O94992, 359 aa) supplies the inhibition in the 7SK snRNP: 7SK binding alone does not inhibit P-TEFb, and HEXIM1 requires 7SK as the adapter that lets it reach cyclin T1 [E1]. HEXIM proteins form dimers through their C-terminal domains and a single HEXIM multimer recruits **multiple** P-TEFb units, so one particle inactivates more than one kinase [E2]. HEXIM1 is self-regulating: releasing P-TEFb induces HEXIM1 transcription from an unannotated proximal promoter, and that induction requires the SEC subunits AFF4 and ELL2 [E3] - the released kinase is spent rebuilding its own cage, through the same scaffold the fusion uses. In KMT2A-rearranged AML this has been tested. In MOLM-13 (KMT2A::MLLT3), shRNA knockdown of HEXIM1 raised c-Myc and in vitro growth and blunted JQ1's differentiation, apoptosis and in vivo survival benefit, while tetracycline-induced ectopic HEXIM1 alone markedly attenuated suspension growth [E4]. Independently, autophagy degrades HEXIM1 (and NPM1) in AML cells expressing NPM1c or MLL-fusion proteins, so the brake pad is actively thinned in these cells [E5].

Go deeper
The answer to 'has anyone raised HEXIM1 in a KMT2A-rearranged cell'

Yes, once, and it worked in vitro. Devaraj and colleagues built MOLM-13 cells (KMT2A::MLLT3, t(9;11) AML) carrying a tetracycline-inducible FLAG-HEXIM1 and found that inducing HEXIM1 alone, without changing CDK9 or cyclin T1 levels, markedly attenuated suspension-culture growth; adding JQ1 then produced more differentiation and more apoptosis than in uninduced cells 4.

The mirror experiment is the stronger one. HEXIM1 knockdown in the same line raised c-Myc, accelerated in vitro growth, and significantly compromised JQ1's survival benefit in NOD/SCID xenografts - so HEXIM1 is not a bystander of BET inhibition but part of its mechanism 4. Notably, HEXIM1 knockdown by itself did not change in vivo growth, which is the honest limit: HEXIM1 is a modifier of drug response here rather than a standalone dependency.

What has never been done: any of this in KMT2A::AFF1, in B-lineage disease, or in infant leukaemia. MOLM-13 is myeloid and carries a different partner.

The brief's premise about HDAC inhibitors is backwards, and the correction matters

It is true that HEXIM1 is induced by HMBA and by HDAC inhibitors - the protein is named for that induction. But the acute effect of HMBA, SAHA and of the BET inhibitor JQ1 is to release free P-TEFb from the 7SK snRNP, not to sequester more of it; HEXIM1 induction is the downstream feedback that reassembles the particle afterwards 673.

So 'give an HDAC inhibitor to raise HEXIM1 and brake elongation' inverts the sequence. The first thing that happens is a transient surge of free P-TEFb. Any intervention built on this axis has to say which phase it is exploiting, and a molecule that raises HEXIM1 without first releasing P-TEFb is a different chemical problem. HMBA analogues optimised specifically as HEXIM1 inducers exist, though they were characterised in LNCaP prostate cells and HMBA's own short half-life is the stated reason it failed clinically 8.

Dependency, and why the number points the other way
armmean Chronos
KMT2A-r, myeloid (n=6)-0.208
KMT2A-r, lymphoid (n=4)-0.267
other leukaemia lines-0.385
all 1,208 lines-0.339

HEXIM1 ranks 17,307th of 17,728 on index-minus-comparator difference (+0.130), i.e. near the anti-selective extreme: losing HEXIM1 costs KMT2A-rearranged lines less than it costs other leukaemia lines, and the direction is the same in both lineage halves 9.

Two readings, and the data cannot separate them. Either KMT2A-rearranged cells already run with more free P-TEFb, so removing the brake adds less - which is the prediction of hyp-kmt2a-aff1-shifts-the-free-versus-sequestered-ptefb-balance; or the difference is 0.13 on four and six lines and is noise. It is also worth saying plainly that HEXIM1 knockout is mildly depleting everywhere (-0.339 across all lines), which is not what one expects of a pure growth suppressor and probably reflects the reservoir's role in stabilising P-TEFb rather than only in inhibiting it.

Expression is not limiting: median 20.2 nTPM, detected in all seven panel lines 10.

The epigenetic column for HEXIM1: not a mark, and the HMBA target turns out to be a histone demethylase

HEXIM1's mechanism is kinase sequestration inside a ribonucleoprotein. It has no catalytic chromatin domain and places no mark, and the honest epigenetic entry starts by refusing a category error this entry is unusually exposed to, because HEXIM1 is pharmacologically coupled to two classic epigenetic drug classes.

The refusal, and the experiment that licenses it. This entry already carries the decisive control as E7 and it deserves to be named as the epigenetic finding: HDAC inhibitors that release free P-TEFb from the 7SK snRNP do so without histone acetylation tracking the outcome - neither histone H3 nor tubulin acetylation correlated with the readout in Jurkat and HeLa cells, and the conclusion was that these compounds work via release of free P-TEFb E11. So pause release is not a chromatin mechanism here, and an entry that wrote "HDAC inhibitor, therefore histone acetylation" would be contradicted by the paper it was citing.

The addition that changes the drug story: the HEXIM1 locus is repressed by an H3K4 demethylase, and that demethylase is the HMBA target. This entry currently records HMBA analogues as HEXIM1 inducers and notes that HMBA failed clinically, with the field's stated objections being poor pharmacological properties, unfavourable ADMET and lack of a known target. The target has since been identified: chemical proteomics, biotin-NeutrAvidin pull-down, surface plasmon resonance and molecular docking converge on KDM5B, the H3K4me3/me2 demethylase, as the target of small-molecule HEXIM1 inducers; RNAi knockdown of KDM5B induces HEXIM1 expression, and chromatin immunoprecipitation supports KDM5B as a direct negative regulator of HEXIM1 E12. That reframes the chemistry: raising HEXIM1 with these compounds is inhibiting a histone demethylase, which is a druggable, well-precedented target class with its own liabilities - and it is a genuinely epigenetic mechanism sitting one step upstream of the brake pad, rather than at it.

The one place HEXIM1 touches an eraser directly. CLP-1/HEXIM1 interacts with MyoD and with histone deacetylases, maximally at the early stage of C2C12 differentiation, and ChIP places a CLP-1/MyoD/HDAC complex on the cyclin D1 promoter, a gene downregulated in differentiated muscle E13. This is the only result found in this tranche in which HEXIM1 is part of a chromatin-bound complex containing a histone-modifying enzyme. It is murine skeletal myogenesis and nothing about leukaemia follows from it, but it establishes that HEXIM1 can sit in such a complex, which is a different statement from the sequestration mechanism the rest of this entry describes. Note the identifier hazard: CLP-1 is logged in data/symbol-collisions.json as an alias of three different genes, one of which owns the symbol; this paper resolves it itself by writing "CLP-1/HEXIM1" throughout, so no inference was needed.

And the chromatin step in the JQ1 arm belongs to BRD4, not to HEXIM1. JQ1 inhibits BRD4 by blocking its association with chromatin, and that releases free P-TEFb from the 7SK snRNP and raises HEXIM1 levels, with the induction requiring AFF4 and ELL2. The chromatin event is upstream and is BRD4's; HEXIM1's part is a transcriptional feedback response.

What has not been done. No KDM5B measurement, and no HEXIM1-locus H3K4me3 measurement, in any KMT2A-rearranged cell was found in this tranche. Given that this entry records the one deliberate HEXIM1-raising experiment in KMT2A-rearranged AML as having slowed growth, a KDM5B inhibitor is now an obvious and previously invisible way to attempt the same thing pharmacologically.

Facts and figures
uniprotO94992
length AA359
gene symbolHEXIM1
uniprot verifiedO94992 = HEXI1_HUMAN, primary gene name HEXIM1 (synonyms CLP1, EDG1, HIS1, MAQ1), 359 aa, Homo sapiens; UniProt REST fetched and read back 2026-09-20, primary gene name asserted
paralogueHEXIM2 (Q96MH2, 286 aa)
rolesupplies the kinase inhibition inside the 7SK snRNP; binds cyclin T1 through its C-terminal domain
breakpoint sidenot applicable - no documented contact with either half of KMT2A; acts on P-TEFb, which the fusion reaches in trans
subnuclear localizationparaspeckle-component association via NEAT1 (HDP-RNP), plus a measured subdiffusive nucleoplasmic state set by 7SK and P-TEFb binding. No speckle, nucleolar, PML-body or heterochromatin assignment.
shown for fusionNO. No HEXIM1 imaging, fractionation or subnuclear assignment has been performed in any KMT2A-rearranged cell. This absence is load-bearing rather than clerical: HEXIM1 is the inhibitory subunit of the 7SK snRNP, the reservoir that every mechanistic account of KMT2A::AFF1 depends on because the fusion's SEC captures P-TEFb released from it. The graph already records (Q62, cycle 36) that CDK9 itself has never been localised in a KMT2A-rearranged cell. So NEITHER SIDE OF THE P-TEFb EQUILIBRIUM - not the kinase, not its inhibitor - has ever been imaged in this disease, and the reservoir model is entirely inferential in spatial terms. Zero in titles and abstracts; supporting figures not searched.
epigenetic readoutPAUSE BIOLOGY IS NOT A CHROMATIN MARK, AND THIS ENTRY SAYS SO EXPLICITLY. HEXIM1 writes, erases and reads no histone modification: it is the inhibitory subunit of the 7SK snRNP and its mechanism is sequestration of a kinase, not modification of a nucleosome. The strongest evidence that the pause axis must not be relabelled as chromatin is already in this entry as E7 and is worth restating as the epigenetic finding it is: HDAC inhibitors that release free P-TEFb from the 7SK snRNP do so WITHOUT histone acetylation tracking the outcome - NEITHER histone H3 NOR tubulin acetylation correlated with the readout in Jurkat and HeLa cells, and the authors concluded the effect works via release of free P-TEFb (PMID 23539624). An HDAC inhibitor acting on this axis is therefore not acting through a histone mark at the step that matters. THREE GENUINE EPIGENETIC CONNECTIONS, none of them a mark HEXIM1 places: (1) ITS OWN LOCUS IS REPRESSED BY AN H3K4 DEMETHYLASE - KDM5B, the H3K4me3/me2 demethylase, directly negatively regulates HEXIM1; RNAi knockdown of KDM5B induces HEXIM1 expression, ChIP supports direct regulation, and chemical proteomics plus biotin-NeutrAvidin pull-down, SPR and docking identify KDM5B as THE TARGET OF THE SMALL-MOLECULE HEXIM1 INDUCERS, i.e. the molecular target HMBA was said to lack (PMID 31805991). (2) IT RECRUITS AN ERASER AT A PROMOTER - CLP-1/HEXIM1 interacts with MyoD and with histone deacetylases maximally early in C2C12 differentiation, and ChIP places a CLP-1/MyoD/HDAC complex on the cyclin D1 promoter (PMID 20940258; murine skeletal muscle, far from this disease, and note CLP-1 is a registry-logged ambiguous alias - this paper resolves it itself by writing 'CLP-1/HEXIM1'). (3) A CHROMATIN READER RELEASES IT - JQ1 inhibits BRD4 by blocking its association with chromatin and thereby releases free P-TEFb from the 7SK snRNP, inducing HEXIM1 transcription, with the induction requiring the SEC subunits AFF4 and ELL2. So the chromatin step is BRD4's, and HEXIM1's response to it is transcriptional. CAUSE OR CONSEQUENCE: for KDM5B the perturbation is genetic and chemical and the direction KDM5B -> HEXIM1 repression is established; NOT in a KMT2A-rearranged cell, and no KDM5B measurement in this disease was found in this tranche.
expression hpa celline nTPM
RS4;119.800RS4;11: 9.8SEM15.2SEM: 15.2NOMO-118.7NOMO-1: 18.7THP-120.2THP-1: 20.2median20.2median: 20.2OCI-AML-221.9OCI-AML-2: 21.9MOLM-1322.1MOLM-13: 22.1MV4-1132.9MV4-11: 32.9
essentiality
{
  "chronos_mean": -0.339,
  "n_lines": 1208,
  "chronos_mean_kmt2ar_myeloid": -0.208,
  "n_myeloid": 6,
  "chronos_mean_kmt2ar_lymphoid": -0.267,
  "n_lymphoid": 4,
  "chronos_mean_other_leukemia": -0.385,
  "scan_rank": 17307,
  "scan_n_genes": 17728,
  "difference": 0.13,
  "interpretation": "anti-selective direction, consistent across the lineage split, magnitude at the edge of noise"
}
expression hpa celline nTPM — the numbers behind the chart
{
  "SEM": 15.2,
  "RS4;11": 9.8,
  "MV4-11": 32.9,
  "THP-1": 20.2,
  "NOMO-1": 18.7,
  "MOLM-13": 22.1,
  "OCI-AML-2": 21.9,
  "median": 20.2
}
tested in kmt2a rearranged
{
  "line": "MOLM-13",
  "fusion": "KMT2A::MLLT3",
  "manipulations": [
    "shRNA knockdown",
    "tetracycline-inducible ectopic FLAG-HEXIM1"
  ],
  "source_pmid": "26148704"
}
chemical inducers
[
  "HMBA (hexamethylene bisacetamide)",
  "SAHA and other HDAC inhibitors",
  "JQ1 and BET bromodomain antagonists",
  "unsymmetrical HMBA analogues (PMID 24503105)"
]
related entries
[
  "protein-7sk-snrnp",
  "gene-hexim2",
  "gene-brd4",
  "gene-cdk9"
]
localization
{
  "subcellular": "Nucleus. UniProt O94992 (HEXI1_HUMAN, Protein HEXIM1, 359 aa) was read back only to confirm the identifier; the nuclear assignment and everything below come from imaging and biochemistry.",
  "subnuclear": "PARASPECKLE-ASSOCIATED, and separately measured as a diffusion state rather than a body. HEXIM1 nucleates a ribonucleoprotein built on the long noncoding RNA NEAT1 - the HDP-RNP - whose components are DNA-PK subunits (DNA-PKcs, Ku70, Ku80) and the PARASPECKLE proteins SFPQ, NONO, PSPC1, RBM14 and MATRIN3; HEXIM1 binding to NEAT1 is required for assembly. NEAT1 is the architectural RNA of paraspeckles, so this places a defined HEXIM1 pool in paraspeckle territory. Independently, fluorescence spectroscopy shows HEXIM1 is SUBDIFFUSIVE in the nucleoplasm and that the anomaly is generated by its interactions with 7SK RNA and with P-TEFb, with the oligomeric mixture setting the diffusion exponent - a physical rather than a compartmental description of where HEXIM1 is.",
  "how_shown": "immunoprecipitation, mass spectrometry and RNA sequencing to define the HDP-RNP (PMID 28712728); fluorescence correlation spectroscopy plus interaction-deficient mutants and numerical simulation for the diffusion measurement (PMID 31590891)",
  "systems": "human non-leukaemic cell lines for both; no KMT2A-rearranged system",
  "node": "localization-paraspeckles-neat1-kmt2a-r"
}
localization counted absence
{
  "query": "HEXIM1[tiab] AND (localiz* OR nuclear OR nucleolar OR speckle* OR cytoplasm* OR paraspeckle*)[tiab]",
  "records": 68,
  "records_in_a_kmt2a_context": 0,
  "operand_controls": {
    "\"7SK\"[tiab] AND (speckle* OR nucleolus OR nucleolar OR \"nuclear bodies\" OR localiz*)[tiab]": 35
  },
  "date": "2026-09-21",
  "tool": "PubMed E-utilities esearch",
  "bound": "zero in titles and abstracts; supporting figures not searched",
  "names_searched": [
    "HEXIM1",
    "7SK",
    "CLP1",
    "EDG1",
    "HIS1",
    "MAQ1"
  ]
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.2533,
  "chronos_mean_other_leukaemia": -0.3829,
  "difference": 0.1296,
  "pan_essentiality_all_lines": -0.3394,
  "selectivity_rank": 17307,
  "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."
}
cell type
{
  "evidence_class_key": "concept-expressed-required-maintains-three-evidence-classes",
  "expressed_in": {
    "class": 1,
    "summary": "broadly expressed; HEXIM1 is the inhibitory subunit of the 7SK snRNP and is present wherever P-TEFb is regulated. This entry already carries HPA cell-line nTPM values in data.expression_hpa_celline_nTPM."
  },
  "required_in": {
    "class": 2,
    "claim": "In HUMAN ERYTHROPOIESIS, HEXIM1 is an essential transcription regulator: HEXIM1 controls erythroid gene expression and function, and overexpression promotes erythroid proliferation and fetal globin expression by enforcing RNA polymerase II pausing at cell-cycle checkpoint genes while increasing RNAPII occupancy at genes that promote cycle progression [E_q69_ct1].",
    "pmids": [
      "37738561"
    ],
    "system_named_in_the_claim": "HUMAN erythroid cells - one of very few results in this tranche that is human and primary-cell rather than murine",
    "why_this_is_the_first_haematopoietic_cell_type_for_HEXIM1_in_this_graph": "the entry previously carried no cell type at all. The erythroid compartment is not this disease's lineage, and that is stated rather than smoothed over - but it is the one blood cell type in which HEXIM1's requirement has actually been measured."
  },
  "not_required_in": {
    "claim": "NO KMT2A-selective requirement, and the placement is extreme. Chronos -0.2533 in the KMT2A-rearranged arm against -0.3829 in other leukaemias, difference +0.1296, rank 17,307 of 17,728 - the deepest ANTI-SELECTIVE placement in this tranche after HDAC7 and CDKN2A. Pan-essentiality -0.3394, so there is dynamic range and the result is interpretable.",
    "what_that_means_mechanistically": "HEXIM1 is the INHIBITOR of P-TEFb. A gene whose loss releases P-TEFb should, if anything, be LESS required in a leukaemia whose mechanism is P-TEFb recruitment - which is the direction the scan reports. The anti-selectivity is consistent with the mechanism rather than in tension with it, and reading it as a failed dependency would be a category error."
  },
  "nodes": [
    "cell-type-erythroid-progenitor",
    "cell-type-mep"
  ],
  "node_update": "2026-09-21, Q73: cell-type-erythroid-progenitor now exists and is the correct target for this claim, which is measured on globin loci with GATA1 redistribution in committed human erythroid cells. cell-type-mep, the BIPOTENT megakaryocyte-erythroid progenitor, was an approximation reported as such by Q69 and is retained as the upstream relation rather than deleted. Edged as `expressed_in` (genome-wide HEXIM1 profiling was performed in human erythroid cells) AND separately as `regulates` (the requirement).",
  "context": "human erythroid cells for the requirement; human epidermal keratinocytes for the state; no measurement in any KMT2A-rearranged cell"
}
cell state
{
  "evidence_class_key": "concept-expressed-required-maintains-three-evidence-classes",
  "maintains_state": {
    "class": 3,
    "states": [
      "cell-state-fetal-program-retention",
      "cell-state-differentiation-block"
    ],
    "claim": "In HUMAN EPIDERMAL PROGENITORS, the AFF1-scaffolded Super Elongation Complex associates with HEXIM1-containing INACTIVE CDK9 to suppress a group of 'rapid response' genes in the progenitor state; switching CDK9 on - with a SEC peptidomimetic inhibitor or PKC signalling - induces those genes within hours and initiates terminal differentiation. The HEXIM1/CDK9 activity switch is what bifurcates progenitor fate between self-renewal and differentiation [E_q69_cs1]. Separately, in HUMAN ERYTHROID cells, HEXIM1 overexpression shifts RNAPII and GATA1 from the beta-globin to the gamma-globin locus and raises FETAL globin and the lncRNA BGLT3 [E_q69_ct1].",
    "pmids": [
      "35906225",
      "37738561"
    ],
    "system_named_in_the_claim": "human keratinocytes for the progenitor switch; human erythroid cells for the fetal-globin state - neither is a leukaemia"
  },
  "why_the_epidermal_result_belongs_in_a_leukaemia_graph": "the scaffold is AFF1. This graph's index fusion is KMT2A::AFF1, and the mechanism it asserts is AFF1-mediated SEC recruitment and P-TEFb activation. PMID 35906225 shows the SAME scaffold, in a completely different tissue, holding CDK9 INACTIVE through HEXIM1 to keep a progenitor undifferentiated. That is a differentiation block built out of AFF1 and HEXIM1 with no fusion anywhere - which is directly relevant to the question of whether this disease's differentiation block is fusion-driven or is a normal progenitor programme the fusion locks in.",
  "bounded_query": {
    "date": "2026-09-21",
    "tool": "NCBI esearch db=pubmed",
    "query": "HEXIM1[ti] AND (hematopoie* OR haematopoie* OR erythroid OR megakaryo* OR myeloid OR differentiation)[tiab]",
    "count": 18,
    "on_topic_after_title_check": 2,
    "names_searched": [
      "HEXIM1",
      "CLP-1",
      "MAQ1",
      "EDG1",
      "HIS1"
    ],
    "alias_warning": "EDG1 is unusable as a HEXIM1 alias in a title/abstract search - it is also the old symbol for the sphingosine-1-phosphate receptor S1PR1. CLP-1 additionally matches a C. elegans gene. Counts run over the alias set were not used.",
    "bound": "titles and abstracts only; supporting figures not searched (brief rule 80)"
  },
  "confidence": "moderate",
  "caveats": "Neither system is haematopoietic malignancy and neither is this lineage. The erythroid result's headline effects come from HEXIM1 OVERexpression, which is a gain-of-function perturbation and weak evidence about the endogenous protein's dose; the paper does establish an essential role but the fetal-globin switch specifically is an overexpression phenotype. The epidermal result is keratinocytes. And the direction of HEXIM1's effect is not fixed - the same paper reports HEXIM1 increased at both repressed AND activated genes, with GATA1 co-occupancy determining the sign, so 'HEXIM1 represses' is not a safe general statement.",
  "counter_evidence": {
    "note": "HEXIM1's sign is context-dependent, which undercuts any simple state assignment. In the erythroid system HEXIM1 is found at both repressed and activated genes and the outcome depends on whether GATA1 is gained or lost at the same locus - genes gaining both HEXIM1 and GATA1 gain RNAPII and expression, genes gaining HEXIM1 but losing GATA1 gain pausing and lose expression. A model in which HEXIM1 simply holds P-TEFb inactive and thereby blocks differentiation does not survive that observation.",
    "citation": {
      "title": "HEXIM1 is an essential transcription regulator during human erythropoiesis.",
      "authors": "Lv X, Murphy K, Murphy Z, Getman M, Rahman N, Nakamura Y, Blanc L, Gallagher PG, Palis J, Mohandas N, Steiner LA",
      "year": 2023,
      "venue": "Blood",
      "pmid": "37738561",
      "doi": "10.1182/blood.2022019495",
      "type": "paper"
    }
  }
}
How this could bite us back
  • Raising HEXIM1 brakes elongation in every proliferating cell, not only in the blast. The MOLM-13 experiment used a tetracycline-inducible transgene inside the leukaemia cell; a drug has no such targeting.
  • HEXIM2 compensates quantitatively for HEXIM1 loss to hold bound P-TEFb constant, and by symmetry a cell can buffer an induced excess. Any HEXIM1-raising agent should be assayed against the bound fraction, not against HEXIM1 protein.
  • HMBA-class HEXIM1 inducers release free P-TEFb acutely before HEXIM1 rises. In a leukaemia whose problem is too much elongation, the first hours of dosing may push the wrong way.
  • HEXIM1 loss is mildly depleting across all 1,208 DepMap lines, which means the reservoir also stabilises P-TEFb. An agent that empties or destroys the particle is not the same as one that fills it.
  • HMBA's clinical failure is attributed to a short half-life, and the optimised analogues have no in vivo or selectivity data at all.
  • neither side of the P-TEFb equilibrium has been imaged in this disease - The reservoir model - fusion SEC captures P-TEFb released from the 7SK snRNP - is the mechanistic basis for CDK9 inhibition in KMT2A-rearranged leukaemia. CDK9 has never been localised in a KMT2A-rearranged cell (recorded by Q62) and neither has HEXIM1. A HEXIM1-inducing agent such as a BET inhibitor or HMBA is predicted to act by shifting a spatial equilibrium that has never been observed in this disease.
  • A KDM5B inhibitor used to raise HEXIM1 would be an H3K4-demethylase inhibitor given to a leukaemia whose defining lesion already carries the KMT2A H3K4 methyltransferase moiety, and whose chromatin shows paradoxical H3K4me3 gains when H3K79me2 is depleted. Raising H3K4 methylation globally in that cell is not an obviously safe way to reach one gene, and the off-HEXIM1 consequences would be the dominant pharmacology rather than a side effect.
Still open
  1. Does inducing HEXIM1 attenuate growth in a KMT2A::AFF1 B-lineage line (SEM, RS4;11) as it does in MOLM-13? The experiment is one lentivirus away and has never been run.
  2. Is HEXIM1 protein lower in KMT2A::AFF1 blasts than in other leukaemia blasts, given that autophagy degrades it in MLL-fusion AML? No blot exists.
  3. Does the HEXIM1 proximal promoter that carries poised, P-TEFb-sensitive Pol II sit inside a KMT2A fusion binding site? If the fusion occupies its own brake's promoter, the feedback loop has a direct oncogenic input.
  4. Is HEXIM1 - or 7SK, or CDK9 - visibly redistributed in a KMT2A-rearranged cell relative to a non-rearranged one? The reservoir model that justifies CDK9 inhibition in this disease makes a spatial prediction that has never been tested on either side of the equilibrium.
  5. Would a KDM5B inhibitor raise HEXIM1 in a KMT2A-rearranged cell, and would it slow growth the way tetracycline-induced ectopic HEXIM1 did in MOLM-13? KDM5B is now the identified molecular target of the HEXIM1-inducer chemical series - the target HMBA was said to lack - which converts 'raise HEXIM1 pharmacologically' from a compound-quality problem into a target-class problem with existing chemical matter. No KDM5B measurement in KMT2A-rearranged leukaemia was found in this tranche.
  6. Does a KDM5B inhibitor trigger the ACUTE P-TEFb release phase before the HEXIM1 induction phase? Every other agent on this axis does, and in a KMT2A-rearranged cell a transient surge of free P-TEFb feeds the fusion's own elongation route. The order of the two phases decides whether this is a brake or an accelerator, and it is unmeasured.
  7. Is the HEXIM1 promoter's H3K4me3 state different in KMT2A-rearranged cells? If KDM5B sets HEXIM1 levels by demethylating H3K4 at that promoter, and if autophagy is separately thinning HEXIM1 protein in MLL-fusion-expressing cells as this entry already records, then the brake pad is being removed at two levels at once - and only the protein level has been looked at.
Every source behind this entry (17)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    HEXIM1 inhibits P-TEFb in a process that specifically requires 7SK to mediate the HEXIM1:P-TEFb interaction; 7SK binding alone is not sufficient to inhibit the kinase.

    PMID 14580347DOI
    What this evidence cannot show

    HeLa biochemistry and in vitro transcription; no haematopoietic cell.

    Evidence that argues against this (1)
    • A reconstituted system shows 7SK can bind P-TEFb independently of HEXIM1, governed by CDK9 T-loop phosphorylation, so HEXIM1 is the inhibitor but not necessarily the first component to engage. Chen R et al. 2004
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    HEXIM1 and HEXIM2 interact through their C-terminal domains and form dimers, and multiple P-TEFb modules are found bound to a single HEXIM1 multimer within the inactive P-TEFb/HEXIM1/7SK complex.

    Transcription-dependent association of multiple positive transcription elongation factor units to a HEXIM multimer. · Dulac C, Michels AA, Fraldi A, Bonnet F, Nguyen VT, Napolitano G, Lania L, Bensaude O · J Biol Chem 2005
    PMID 15994294DOI
    What this evidence cannot show

    Hydrodynamic measurements, gel filtration and chemical cross-linking on recombinant and cellular protein; 'dimers' is inferred from cross-linking and anisotropy rather than from a structure. No structure of the assembled particle exists.

    Evidence that argues against this (1)
    • The compensation experiments show HEXIM1 and HEXIM2 can substitute quantitatively for each other in the bound fraction, which means the composition of any given particle in vivo is not fixed by these measurements. Yik JH et al. 2005
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Release of P-TEFb from the 7SK snRNP increases synthesis of HEXIM1 but not HEXIM2, from an unannotated proximal promoter carrying P-TEFb-sensitive poised Pol II, and the induction requires recruitment of the super elongation complex subunits AFF4 and ELL2.

    PMID 24515107DOI
    What this evidence cannot show

    HeLa cells, ChIP-seq plus luciferase reporters. The promoter's use in haematopoietic cells was not tested, and the feedback was driven with P-TEFb-releasing compounds rather than a physiological signal.

    Evidence that argues against this (1)
    • HEXIM2 does not participate in this feedback, so a system described as self-correcting is self-correcting only through one of its two paralogues, and a cell that leans on HEXIM2 would not show it. Byers SA et al. 2005
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In MOLM-13 (KMT2A::MLLT3) AML cells, shRNA knockdown of HEXIM1 raised c-Myc and in vitro growth, reduced JQ1-induced differentiation and apoptosis, and significantly compromised JQ1's survival benefit in NOD/SCID xenografts, while tetracycline-induced ectopic FLAG-HEXIM1 alone markedly attenuated suspension-culture growth without changing CDK9 or cyclin T1 levels.

    HEXIM1 induction is mechanistically involved in mediating anti-AML activity of BET protein bromodomain antagonist. · Devaraj SG, Fiskus W, Shah B, Qi J, Sun B, Iyer SP, Sharma S, Bradner JE, Bhalla KN · Leukemia 2016
    PMID 26148704DOI
    What this evidence cannot show

    One KMT2A-rearranged line (MOLM-13, a KMT2A::MLLT3 AML) plus one patient-derived AML sample whose genotype is not stated in the text. Not KMT2A::AFF1, not B-lineage, not infant. The PubMed record carries no abstract; this is read from the PMC full text (PMC4809433) reached through efetch db=pmc after fullTextXML returned HTTP 500. HEXIM1 knockdown alone had little effect on in vivo growth, so HEXIM1 modifies drug response rather than being a standalone dependency.

    Evidence that argues against this (1)
    • The authors themselves note that BRD4's ET domain can activate transcription independently of P-TEFb by recruiting JMJD6 or NSD3, which is why HEXIM1 knockdown does not abolish the BET-inhibitor response; the HEXIM1 arm is one route among several. Devaraj SG et al. 2016
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Autophagy is activated in acute leukaemia cells expressing mutant NPM1 or MLL-fusion proteins and degrades NPM1 and HEXIM1, two negative regulators of the BET pathway; blocking autophagy pharmacologically or by ATG5 knockdown raises both proteins.

    Autophagy mediates proteolysis of NPM1 and HEXIM1 and sensitivity to BET inhibition in AML cells. · Huang M, Garcia JS, Thomas D, Zhu L, Nguyen LX, Chan SM, Majeti R, Medeiros BC, Mitchell BS · Oncotarget 2016
    PMID 27732946DOI
    What this evidence cannot show

    AML cell lines and primary AML; the MLL-fusion arm is not resolved to a partner and no KMT2A::AFF1 or B-lineage cell appears. The causal chain from HEXIM1 loss to BET-pathway activation is inferred from correlated protein levels rather than demonstrated by rescue.

    Evidence that argues against this (1)
    • JQ1 and I-BET-151 themselves inhibit autophagy and raise HEXIM1 in the same system, so the drug and the proposed mechanism move the same marker, and the direction of causality between autophagy inhibition and BET response is not separated. Huang M et al. 2016
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Hexamethylene bisacetamide, SAHA and other histone deacetylase inhibitors cause a rapid release of free P-TEFb from the 7SK snRNP followed by its reassembly, and the BET inhibitor JQ1 does the same, transiently raising free P-TEFb and the Brd4/P-TEFb and SEC/P-TEFb complexes, with HEXIM1 induction as a downstream consequence.

    PMID 22952229DOI
    What this evidence cannot show

    HeLa and Jurkat cells in an HIV-latency framework; the readout for 'activation' is HIV transcription, not a leukaemia programme.

    Evidence that argues against this (1)
    • Effects of the same compounds in a cell where P-TEFb levels are very low are minimal, so the size and even the existence of the release depends on how much P-TEFb the cell has - which is exactly the quantity nobody has measured in a KMT2A-rearranged blast. Bartholomeeusen K et al. 2013
  7. E7established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    HDAC inhibitors that reactivate HIV do so by releasing free P-TEFb from the inhibitory 7SK snRNP, and neither histone H3 nor tubulin acetylation correlates with the reactivation; in resting primary CD4+ T cells, where P-TEFb levels are very low, SAHA has minimal effect until P-TEFb levels are first raised.

    PMID 23539624DOI
    What this evidence cannot show

    Jurkat, HeLa and primary CD4+ T cells in an HIV-latency model. Nothing here is a leukaemia experiment, and the conclusion should not be transferred to AML blasts without checking the P-TEFb pool size.

    Evidence that argues against this (1)
    • In AML cells the same class of agent is reported to act through HEXIM1 induction, i.e. through the reassembly phase rather than the release phase; the two literatures describe opposite halves of the same cycle and have not been reconciled in one cell type. Devaraj SG et al. 2016
  8. E8moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    HMBA analogues with unsymmetrical structures and increased hydrophobicity were optimised as more potent HEXIM1 inducers, motivated explicitly by HMBA's short half-life having limited its clinical application.

    Lead optimization of HMBA to develop potent HEXIM1 inducers. · Zhong B, Lama R, Ketchart W, Montano MM, Su B · Bioorg Med Chem Lett 2014
    PMID 24503105DOI
    What this evidence cannot show

    A medicinal-chemistry letter with potency measured in LNCaP prostate cancer cells by HEXIM1 induction alone. No leukaemia cell, no pharmacokinetics in vivo, no selectivity data. A ChEMBL-style activity count here would measure how much chemistry was done, not tractability.

    Evidence that argues against this (1)
    • If the acute action of HMBA-class compounds is to release P-TEFb rather than to sequester it, then optimising for HEXIM1 induction may be optimising for the feedback response rather than for the desired pharmacology. Bartholomeeusen K et al. 2012
  9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    HEXIM1 has mean Chronos -0.339 across 1,208 DepMap lines; within the KMT2A-rearranged panel the myeloid half means -0.208 and the lymphoid half -0.267 against -0.385 in other leukaemia lines, and it ranks 17,307th of 17,728 on index-minus-comparator difference (+0.130).

    What this evidence cannot show

    Six myeloid and four lymphoid lines; +0.130 sits at the edge of the +/-0.15 noise band this project uses at that n, so the direction is consistent across the lineage split but the magnitude is not established. A CRISPR knockout of a brake is also a poor model of a pharmacological brake release, because it removes the protein permanently rather than changing a partition.

    Evidence that argues against this (1)
    • In MOLM-13, which is in the KMT2A-rearranged arm, HEXIM1 knockdown measurably increased in vitro growth and blunted a drug response, so a near-zero CRISPR difference does not mean HEXIM1 is functionally inert in these cells. Devaraj SG et al. 2016
  10. E10established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    HEXIM1 is detected in all seven KMT2A-rearranged panel lines the Human Protein Atlas carries, median 20.2 nTPM, so its low dependency score is not an expression artefact.

    What this evidence cannot show

    Bulk poly-A cell-line RNA, not protein and not patient blasts.

    Evidence that argues against this (1)
    • HEXIM1 protein is degraded by autophagy in MLL-fusion AML cells, so mRNA abundance may overstate how much brake pad is actually present. Huang M et al. 2016
  11. E_q65_loc1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    HEXIM1 assembles a ribonucleoprotein on the long noncoding RNA NEAT1 - the HEXIM1-DNA-PK-paraspeckle components-ribonucleoprotein complex - containing DNA-PKcs, Ku70, Ku80 and the paraspeckle proteins SFPQ, NONO, PSPC1, RBM14 and MATRIN3, with HEXIM1 binding to NEAT1 required for assembly; the complex regulates the cGAS-STING-IRF3 response to foreign DNA.

    PMID 28712728DOI
    What this evidence cannot show

    The complex is defined biochemically (immunoprecipitation, mass spectrometry, RNA-seq); the paper is not an imaging study and does not itself show HEXIM1 inside a microscopically resolved paraspeckle. NEAT1 is the paraspeckle scaffold and the co-purifying proteins are paraspeckle proteins, so the compartment assignment is an inference from composition, and this entry says so rather than asserting a visualised address. Human non-leukaemic lines. Abstract only.

    Evidence that argues against this (1)
    • The competing, and better-measured, description of HEXIM1's nuclear whereabouts is not a body at all: fluorescence spectroscopy finds HEXIM1 subdiffusive throughout the nucleoplasm with the anomaly generated by 7SK and P-TEFb binding, which describes a distributed interaction landscape rather than residence in a compartment. Furlan A et al. 2019
  12. E_q65_loc2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    HEXIM1 diffuses anomalously (subdiffusively) in the nucleus, and interaction-deficient mutants show the anomaly is generated by its binding to both 7SK RNA and P-TEFb, with the proportions of distinct oligomeric HEXIM1 subpopulations setting the apparent anomaly parameter of the whole population.

    PMID 31590891DOI
    What this evidence cannot show

    A biophysical measurement in cultured non-leukaemic cells; it constrains where HEXIM1 spends time without naming a compartment, and small changes in oligomer proportions shift the readout, so it is sensitive to expression level. Abstract only.

    Evidence that argues against this (1)
    • A diffusion measurement cannot exclude a resident pool: the HDP-RNP work places a defined, NEAT1-dependent HEXIM1 assembly with paraspeckle proteins, and a minority bound population can be invisible to a population-averaged diffusion exponent. Morchikh M et al. 2017
  13. E_q69_ct1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In HUMAN erythropoiesis, HEXIM1 is an essential transcription regulator: HEXIM1 overexpression promotes erythroid proliferation and fetal globin expression, regulates erythroid proliferation by enforcing RNA polymerase II pausing at cell-cycle checkpoint genes and increasing RNAPII occupancy at genes promoting cycle progression, and produces genome-wide redistribution of GATA1 and RNAPII with loss of both at beta-globin and gain at gamma-globin, raising fetal globin and the lncRNA BGLT3; GATA1 is the determinant of whether HEXIM1 represses or activates a gene.

    HEXIM1 is an essential transcription regulator during human erythropoiesis. · Lv X, Murphy K, Murphy Z, Getman M, Rahman N, Nakamura Y, Blanc L, Gallagher PG, Palis J, Mohandas N, Steiner LA · Blood 2023
    PMID 37738561DOI
    What this evidence cannot show

    Abstract read in full; full text not read. Human erythroid cells - the right species, the wrong lineage for this graph's disease, and not a leukaemia. The fetal-globin and proliferation phenotypes are OVERexpression results; the essentiality claim is separate from them. No KMT2A-rearranged cell appears anywhere in the study.

    Evidence that argues against this (1)
    • The AFF1/HEXIM1 result in epidermal progenitors assigns HEXIM1 the opposite role - holding CDK9 inactive to SUPPRESS a differentiation-initiating gene set, with activation of CDK9 driving differentiation. Whether HEXIM1 promotes or restrains a differentiation programme is therefore tissue-dependent, and this entry does not generalise either result. Lloyd SM et al. 2022
  14. E_q69_cs1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In HUMAN EPIDERMAL PROGENITORS, the Super Elongation Complex suppresses a group of 'rapid response' genes that carry high paused Pol II in the progenitor state and robust Pol II elongation on differentiation; the repression depends on the AFF1 scaffold and not AFF4, AFF1-SEC associates with HEXIM1-containing inactive CDK9 in the progenitor state, and SEC peptidomimetic inhibitors or PKC signalling activate CDK9 and induce the differentiation-activating transcription factors GRHL3, OVOL1, PRDM1 and ZNF750 within hours, so the activity switch of SEC-associated CDK9 underlies the bifurcation of progenitor fate between self-renewal and differentiation.

    CDK9 activity switch associated with AFF1 and HEXIM1 controls differentiation initiation from epidermal progenitors. · Lloyd SM, Leon DB, Brady MO, Rodriguez D, McReynolds MP, Kweon J, Neely AE, Blumensaadt LA, Ho PJ, Bao X · Nat Commun 2022
    PMID 35906225DOI
    What this evidence cannot show

    Abstract read in full; full text not read. Human keratinocytes - an epithelial progenitor system, not haematopoietic and not leukaemic (brief rule 73). Its relevance here is that the scaffold is AFF1, the 5' partner of this graph's index fusion, and that the mechanism is a HEXIM1-dependent differentiation block with no fusion present. That is an analogy and a falsification target, not evidence about KMT2A::AFF1.

    Evidence that argues against this (1)
    • The erythroid study finds HEXIM1 at activated as well as repressed genes and shows the sign is set by GATA1 co-occupancy, so the clean 'HEXIM1 holds CDK9 off and blocks differentiation' model from keratinocytes does not transfer as a general rule. In blood the determinant may be a lineage transcription factor rather than the HEXIM1/CDK9 switch itself. Lv X et al. 2023
  15. E11established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    THE GUARD AGAINST RELABELLING PAUSE RELEASE AS A CHROMATIN MARK: HDAC inhibitors that reactivate HIV do so by releasing free P-TEFb from the inhibitory 7SK snRNP, and neither histone H3 nor tubulin acetylation correlated with reactivation in Jurkat and HeLa cells; in resting primary CD4+ T cells where P-TEFb levels are very low, SAHA had minimal effect until P-TEFb levels were raised with bryostatin 1.

    PMID 23539624DOI
    What this evidence cannot show

    Abstract read from the eutils record; already cited on this entry as E7, and this item exists to state its epigenetic significance rather than to add a new fact - the reader of an epigenetic_readout field needs the negative control in front of them, not two fields away. The system is HIV latency in Jurkat, HeLa and primary CD4+ T cells, NOT leukaemia and NOT a KMT2A-fusion cell. 'Neither histone H3 nor tubulin acetylation correlated' is a bulk correlation across compounds, not a locus-resolved acetylation measurement, so it excludes histone acetylation as the explanatory variable for these compounds' P-TEFb effect without excluding acetylation from having other roles.

    Evidence that argues against this (1)
    • The non-chromatin reading must not be over-extended, because one compound class on this axis does start at chromatin. JQ1 releases free P-TEFb from the 7SK snRNP by inhibiting BRD4 specifically through blocking its association with chromatin, transiently raising free P-TEFb and BRD4-P-TEFb and SEC-P-TEFb complexes and inducing HEXIM1, with ELL2 knockdown blocking the effect. So 'the 7SK reservoir is not reached through chromatin' is false as a general claim: it is reached through a chromatin READER in the BET case and by a chromatin-independent route in the HDAC-inhibitor case, and an entry should say which compound it means. Bartholomeeusen K et al. 2012
  16. E12moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    THE HEXIM1 LOCUS IS REPRESSED BY AN H3K4 DEMETHYLASE, AND THAT DEMETHYLASE IS THE HEXIM1-INDUCER TARGET: KDM5B, the H3K4me3/me2 demethylase, was identified as the target of small-molecule HEXIM1 inducers by chemical proteomics, biotin-NeutrAvidin pull-down, surface plasmon resonance and molecular docking; RNAi knockdown of KDM5B induced HEXIM1 expression, chromatin immunoprecipitation supported direct negative regulation of HEXIM1 by KDM5B, and KDM5B inhibitors altered breast cancer cell differentiation and proliferation in a manner partly dependent on HEXIM1 by shRNA depletion.

    PMID 31805991DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read in this tranche. BREAST CANCER cells and a mouse metastasis model - NOT leukaemia, NOT haematopoietic, and NOT a KMT2A-fusion system, so rule 73 applies to every downstream use. Target identification by chemical proteomics plus SPR plus docking is a strong convergent argument but is not a co-crystal or a cellular occupancy measurement; 'the target' of a chemical series can be one of several engaged proteins. The HEXIM1-dependence of the phenotype is partial by the authors' own design (shRNA depletion of HEXIM1 'followed by' the assays), so KDM5B inhibition does other things too. Not cited anywhere in this corpus before 2026-09-21, verified with scripts/check-citation-reuse.py.

    Evidence that argues against this (1)
    • The direction of the therapeutic idea is the problem, not the target identification. This entry's own corrected framing is that the ACUTE effect of HMBA, SAHA and JQ1 on this axis is to RELEASE free P-TEFb, with HEXIM1 induction arriving afterwards as feedback that reassembles the particle. A KDM5B inhibitor that raises HEXIM1 transcription would be exploiting the feedback phase, and nothing read here establishes that it does not also trigger the release phase first. A compound that transiently increases free P-TEFb in a KMT2A-rearranged cell - where the SEC is the fusion's own elongation route - could accelerate the disease before it brakes it, and no experiment addresses the order. Bartholomeeusen K et al. 2012
  17. E13weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    THE ONE PLACE HEXIM1 SITS IN A CHROMATIN-BOUND COMPLEX WITH A HISTONE-MODIFYING ENZYME: CLP-1/HEXIM1 interacts with both MyoD and histone deacetylases, maximally at the early stage of C2C12 skeletal muscle differentiation, and chromatin immunoprecipitation shows a CLP-1/MyoD/HDAC complex bound to the promoter of the cyclin D1 gene, which is downregulated in differentiated muscle cells; C2C12 CLP-1 heterozygous knockdown cells fail to differentiate in low serum.

    CLP-1 associates with MyoD and HDAC to restore skeletal muscle cell regeneration. · Galatioto J, Mascareno E, Siddiqui MA · J Cell Sci 2010
    PMID 20940258DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. Marked WEAK: murine C2C12 skeletal myoblasts, a system with no relationship to leukaemia, and the paper is a 2010 co-IP-plus-ChIP study without the perturbation that would show the HDAC association is functionally required at that promoter. 'Histone deacetylases' is unresolved to a family member or isoform, so no specific eraser can be named. IDENTIFIER NOTE: CLP-1 is logged in data/symbol-collisions.json as an alias of three different genes - CLP1 (Q92989) owns the symbol, HEXIM1 (O94992) and COLEC12 (Q5KU26) carry it as a synonym - and the registry blocks autogeneration on it. This paper is unambiguous because it writes 'CLP-1/HEXIM1' explicitly; a paper using CLP-1 alone could not be attributed this way. Not cited anywhere in this corpus before 2026-09-21.

    Evidence that argues against this (1)
    • This is the one result that could be read as putting HEXIM1 into histone-modification biology, and the entry's own controlling experiment argues against generalising it. On the axis that matters for this project - P-TEFb sequestration and release - histone acetylation was measured and did NOT correlate with the functional outcome. A promoter-bound HEXIM1/MyoD/HDAC complex in differentiating myoblasts is therefore best read as a distinct, tissue-specific co-repressor role rather than as evidence that HEXIM1's 7SK function is acetylation-coupled. Bartholomeeusen K et al. 2013
Connected to