CCNT1 and CCNT2 have no histone substrate and reach chromatin on someone else's back: Brd4 recruits cyclin T1 and Cdk9 to acetylated chromatin and drives Pol II Ser2 phosphorylation - a borrowed address, which is a third annotation state the corpus has no word for
also called cyclin T1 cyclin T2 chromatin address · q181 CCNT1 CCNT2
Cyclin T1 and T2 are the partner halves of the kinase that switches paused polymerase into productive elongation. They do not touch histones themselves - a BET protein reading acetylated histones brings them to the right place on the genome.
The same thing, for a scientist
gene-ccnt1 and gene-ccnt2 are in this beat's eight-entry real-gap floor. CCNT1 is named in 57 node files of this corpus - more than any other floor member - and carries no epigenetics annotation. The literature gives a clear answer with an unusual shape: Cheung and colleagues showed that Brd4 temporally controls Pol II processivity during transcription elongation THROUGH cyclin T1 and Cdk9 recruitment and Pol II Ser2 phosphorylation. So the cyclin's chromatin address is Brd4's, and Brd4's is acetylated chromatin. The same paper separates Brd2's role: Brd2 associates with CTCF and the cohesin complex to support cis-regulatory enhancer assembly, and binds Stat3 in an acetylation-sensitive manner at active enhancers - which is also the citation this corpus's own gene-brd2, another epigenetics-empty entry, is missing. The annotation state this implies is neither RECORDED nor NOT APPLICABLE nor BOUNDED ABSENCE: it is a BORROWED address, where the protein's chromatin localisation is entirely a property of a named partner. The corpus has no word for it, and it is probably the commonest situation among non-catalytic subunits.
What was shown, and the Brd2/Brd4 split
One paper, two BET proteins, two different jobs at genes they co-regulate 1:
| mechanism | |
|---|---|
| Brd4 | temporally controls Pol II processivity during elongation through cyclin T1 and Cdk9 recruitment and Pol II Ser2 phosphorylation |
| Brd2 | associates with CTCF and the cohesin complex to support cis-regulatory enhancer assembly; binds Stat3 in an acetylation-sensitive manner and facilitates its recruitment to active enhancers occupied by Irf4 and Batf |
The system is mouse Th17 differentiation.
For the cyclins, the operative clause is 'through cyclin T1 and Cdk9 recruitment'. The cyclin does not read a mark; it is delivered to marked chromatin by a reader. CCNT2 is not addressed in this paper at all and its annotation here is by family inference, which is weaker and is labelled as such.
A fourth state: the borrowed address
This beat has argued for three states -- RECORDED, BOUNDED ABSENCE, NOT APPLICABLE [hyp-q181-a-two-state-coverage-metric-cannot-express-not-applicable]. CCNT1 fits none of them cleanly:
- Not RECORDED: no mark is written, read or erased by cyclin T1, and no cyclin T1 ChIP at a defined mark was found.
- Not NOT APPLICABLE: cyclin T1 is demonstrably delivered to chromatin and its kinase partner phosphorylates the polymerase there.
- Not quite BOUNDED ABSENCE: a measurement is not merely missing -- the mechanism is known, and it is that the address belongs to Brd4.
The honest field is a pointer: 'chromatin address borrowed from BRD4, via acetyl-lysine reading; no intrinsic mark activity'. That is a different and more useful annotation than any of the three, and it is probably the right one for a large class of non-catalytic complex subunits -- which, if so, means the three-state proposal is still too small. This is a weakness in this beat's own recommendation, found by trying to apply it.
A correction owed to another beat's entry, reported and not applied
gene-brd2 is in this beat's empty stratum -- no epigenetics field, no mark, and its only chromatin sentence is about BRD4 rather than BRD2 (ana-q181-the-mark-regex-has-no-token-for-enhancer-nucleosome-or-accessibility 2). Its title records it as "a bounded absence: zero KMT2A papers name it".
That bounded absence is correct as stated and incomplete as an epigenetics record. PMID 28262505 gives BRD2 a specific, non-KMT2A chromatin mechanism -- CTCF and cohesin association, acetylation-sensitive Stat3 binding at active enhancers -- that is distinct from BRD4's and that this corpus does not hold anywhere.
Per the standing rule on errors in files I do not own, this is reported, not edited:
- Entry:
gene-brd2 - What is missing: any epigenetics record; the entry's only chromatin content describes BRD4.
- What it should add: BRD2's distinct genomic function -- CTCF/cohesin association supporting enhancer assembly, and acetylation-sensitive Stat3 recruitment -- explicitly scoped as mouse Th17, NOT KMT2A, so the existing bounded absence on KMT2A papers survives intact.
- Citation: PMID 28262505, DOI 10.1016/j.molcel.2016.12.022, abstract read, full text not retrieved.
| epigenetic readout | BORROWED, not intrinsic. Cyclin T1 has no histone substrate and no reader domain; Brd4 recruits cyclin T1 and Cdk9 to acetylated chromatin, where P-TEFb phosphorylates Pol II CTD Ser2 (PMID 28262505). In a KMT2A::AFF1 context the AFF1/AFF4 SEC scaffold is a second and probably dominant delivery route, so the address should be recorded as reader-dependent rather than as Brd4's alone. |
|---|---|
| epigenetic bounded absence | No cyclin T1 or cyclin T2 ChIP, CUT&RUN or histone-mark measurement in any system was found in the searches behind this entry, and nothing in any KMT2A-rearranged cell. CCNT2 is not addressed by the cited paper at all; its annotation here is family inference. |
| epigenetics not applicable | false |
| annotation state proposed | BORROWED ADDRESS - a fourth state, for a protein whose chromatin localisation is entirely a property of a named partner. Distinct from RECORDED, BOUNDED ABSENCE and NOT APPLICABLE, and probably the commonest situation among non-catalytic complex subunits. |
| shown for fusion | NO for CCNT1 and CCNT2 - nobody has looked in any KMT2A fusion. The Brd4-cyclin T1 result is mouse Th17. |
marks named
[]
uniprot
{
"CCNT1": "O60563",
"CCNT2": "O60583"
}correction reported not applied
{
"entry": "gene-brd2",
"problem": "no epigenetics record of any kind; the entry's only chromatin sentence describes BRD4, not BRD2",
"should_add": "BRD2's distinct genomic function - CTCF and cohesin association supporting cis-regulatory enhancer assembly, and acetylation-sensitive Stat3 binding at active enhancers - scoped explicitly as mouse Th17 and NOT KMT2A, so the entry's existing bounded absence on KMT2A papers survives",
"citation": "PMID 28262505, DOI 10.1016/j.molcel.2016.12.022",
"read_depth": "abstract only; full text not retrieved"
}measurements
[
{
"quantity": "node files naming CCNT1 with no entry-level epigenetics record",
"value": 57,
"units": "node files",
"subject": "word-boundary naming count for CCNT1, against gene-ccnt1 carrying no epigenetics field",
"universe": "3,032 node files at commit ac2aeab, via scripts/beat-inventory.py 2026-09-24",
"null": null,
"note": "the most-discussed member of the eight-entry real-gap floor by a wide margin; CCNT2 is named in 17"
}
]- BET inhibitors and menin inhibitors are both in use in KMT2A-rearranged leukaemia and both act on P-TEFb delivery, by different routes. If BRD4 and the SEC scaffold are partly redundant deliverers of cyclin T1/CDK9 to chromatin, then blocking one raises the relative importance of the other - which predicts both a combination rationale and a resistance route, and neither has been measured. The borrowed address is the mechanism that makes the two drugs interact.
- CCNT1 is broadly essential and is the cyclin for a general elongation kinase. Any intervention aimed at cyclin T1 rather than at its delivery is an essential-function toxicity problem, not a selectivity problem, and this corpus's DepMap reasoning should not be read as identifying it as a target simply because it is a dependency.
- Recording CCNT1's epigenetics as 'BRD4 brings it' in a graph where BRD4 is a drug target could invite the inference that hitting BRD4 removes cyclin T1 from chromatin in this disease. In a KMT2A::AFF1 blast the fusion is itself a delivery route, so that inference would be wrong in exactly the setting the graph is for.
- Does cyclin T1 occupy chromatin independently of a BET reader? A cyclin T1 ChIP-seq with and without BET inhibition would settle the borrowed-address reading, and no such dataset was found.
- In a KMT2A::AFF1 blast, what fraction of chromatin-associated cyclin T1 arrives via the fusion's SEC scaffold and what fraction via BRD4? Both routes are documented separately and nobody has partitioned them - which is the experiment that would make the borrowed-address annotation quantitative in the index disease.
- Does CCNT2 behave like CCNT1 at all? Everything in this entry about CCNT2 is family inference from a paper that does not mention it.
- Should the corpus adopt a BORROWED ADDRESS state alongside the three this beat recommends? If it is the commonest case among non-catalytic subunits, the three-state proposal in hyp-q181 is undersized, and that is a defect in this beat's own recommendation.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In mouse Th17 differentiation, Brd4 temporally controls RNA polymerase II processivity during transcription elongation through cyclin T1 and Cdk9 recruitment and Pol II Ser2 phosphorylation, while Brd2 separately associates with the chromatin insulator CTCF and the cohesin complex to support cis-regulatory enhancer assembly and binds Stat3 in an acetylation-sensitive manner, facilitating Stat3 recruitment to active enhancers occupied by Irf4 and Batf.
Distinct Roles of Brd2 and Brd4 in Potentiating the Transcriptional Program for Th17 Cell Differentiation · Cheung KL, Zhang F, Jaganathan A, Sharma R, Zhang Q, Konuma T, Shen T, Lee JY, Ren C, Chen CH, Lu G, Olson MR, Zhang W, Kaplan MH, Littman DR, Walsh MJ, Xiong H, Zeng L, Zhou MM · Molecular Cell 65(6):1068-1080.e5 2017What this evidence cannot show
Abstract-level read; the full text was not retrieved. Mouse Th17 cells - a CD4 T-cell differentiation system, not leukaemia and not any KMT2A-rearranged line. expected_absent_terms includes CCNT2 deliberately: this paper addresses cyclin T1 only, and every statement about CCNT2 in this entry is family inference, not evidence. The 'through cyclin T1 and Cdk9 recruitment' clause is the paper's mechanistic summary; whether Brd4 is the sole or merely a major route by which cyclin T1 reaches chromatin is not addressed.
Evidence that argues against this (2)
- The borrowed-address framing over-reads one system. Brd4 is not the only P-TEFb delivery route - the super elongation complex delivers P-TEFb through AFF1/AFF4 scaffolds, which is the mechanism this entire corpus is built on, and in a KMT2A::AFF1 blast the fusion itself is a delivery route. Saying cyclin T1's address 'is Brd4's' is true in Th17 cells and is likely to be wrong, or at best partial, in the index disease. The annotation this entry proposes should name the route as one of several.
- No search was run for a paper placing cyclin T1 at chromatin independently of a reader, nor for cyclin T1 ChIP-seq in any system; the PubMed query used was a single conjunctive search returning one record, and the bare-operand controls on CCNT1 and CCNT2 were not run. A cyclin T1 ChIP showing occupancy uncorrelated with BET occupancy would refute the borrowed-address reading outright, and nothing rules it out.
supplies the epigenetics record gene-ccnt1 lacks: a borrowed, reader-dependent chromatin address
same proposed annotation for CCNT2, by family inference only - the cited paper does not mention it
Brd4 recruits cyclin T1 and Cdk9 to acetylated chromatin; the cyclin's address is the reader's
reports the missing BRD2-specific epigenetics content (CTCF/cohesin, acetylation-sensitive Stat3) as a correction, not applied
a BORROWED ADDRESS fits none of the three proposed states, so the recommendation is undersized