Oncofusion

start / graph / mechanism-q181-mepce-stops-being-an-enzyme-once-it-is-inside-the-particle

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q181-epigenetics-dimension

MEPCE caps 7SK before assembly and then stops being an enzyme inside the particle - so its epigenetics annotation is a methyltransferase that methylates an RNA phosphate, which is exactly the case a mark-name regex mislabels in both directions

also called MEPCE BCDIN3 readout · q181 MEPCE

MEPCE puts a protective chemical cap on the 7SK RNA. Once the 7SK particle is fully assembled, MEPCE's catalytic site is blocked and it carries on working as a structural piece rather than as an enzyme.

The same thing, for a scientist

gene-mepce is in this beat's eight-entry real-gap floor. Its biology is a clean and instructive case for the coverage metric. MEPCE is a methyltransferase - the methylphosphate capping enzyme for 7SK - and its data payload would therefore trip check-pi-dimensions.py's _EPI_KEY on any key containing 'methylat', scoring as epigenetics-covered on an activity directed at an RNA 5' gamma-phosphate rather than at a histone. Simultaneously, its real chromatin relevance is invisible to the mark regex because it runs through the particle: MEPCE caps 7SK as a LARP7-free enzyme prior to sequestration, loses capping activity once LARP7 binds within the snRNP, and then acts capping-independently to promote the LARP7-7SK interaction, the two cooperating to stabilise 7SK. So the same entry can be over-credited by the key test and under-credited by the mark test, for opposite reasons. The honest annotation is: a recorded RNA-methyltransferase activity that is NOT a histone mark, plus a bounded absence on chromatin.

Go deeper
The biology, in the order it happens
  1. MEPCE caps 7SK, as a LARP7-free enzyme and probably co-transcriptionally, before the RNA is sequestered into the snRNP 1.
  2. LARP7 binds and MEPCE stops catalysing. Within the assembled 7SK snRNP MEPCE loses capping activity, probably because LARP7 occludes its catalytic centre 1.
  3. MEPCE keeps working as a structure. Capping-independently, it promotes the LARP7-7SK interaction, and the two act cooperatively to stabilise 7SK and maintain the integrity of the particle 1.

Step 2 is the unusual part and it is the reason a per-protein annotation is hard. MEPCE's enzymatic activity and its steady-state function are in different phases of the same protein's life, so a field saying 'methyltransferase' describes only the first.

Why this entry is a test case for the metric, in both directions

This beat's two defect findings both bite on MEPCE, and they bite in opposite directions:

  • _EPI_KEY over-credits. The regex matches methylat as a bare substring of a key name. Any author writing methyltransferase_activity on gene-mepce would score the entry as epigenetics-covered [ana-q181-the-epigenetics-detector-reads-key-names-and-never-field-values]. The substrate is an RNA 5' gamma-phosphate. This is the same shape as Rule 89's defect -- a metric counting a link near the thing instead of the thing.
  • _EPI_MARK under-credits. MEPCE's genuine chromatin relevance is the enhancer occupancy of the particle it stabilises [mechanism-q181-7sk-occupies-enhancers-and-recruits-baf-so-the-brake-has-a-chromatin-address], which names no histone residue and is invisible to the mark regex [ana-q181-the-mark-regex-has-no-token-for-enhancer-nucleosome-or-accessibility].

An entry can therefore be wrongly covered and wrongly uncovered at once, and the two errors do not cancel -- they land on different lines of the same report. This is the strongest single argument in the beat for reading values and not key names.

What this entry does not claim
  • No histone methyltransferase activity for MEPCE, in any direction. The searches behind this entry found no report of one and none is asserted. Recording this explicitly matters because the symbol is a methyltransferase and the family assumption is the obvious error.
  • Nothing in a KMT2A-rearranged system. The evidence is HeLa biochemistry from 2009.
  • No claim about MEPCE's own chromatin occupancy. The address belongs to the RNA; assigning it to MEPCE is an inference and is labelled as one in data.epigenetic_bounded_absence.
Facts and figures
epigenetic readoutMethylphosphate capping of the 7SK small nuclear RNA 5' gamma-phosphate, followed by a capping-INDEPENDENT structural role stabilising the LARP7-7SK interaction (PMID 19906723). The methyltransferase activity is directed at an RNA, not at a histone.
epigenetic bounded absenceNo histone-methyltransferase activity, histone-mark measurement, chromatin occupancy or accessibility measurement for MEPCE was found in any system, and none in any KMT2A-rearranged cell. MEPCE's own chromatin occupancy has never been assayed; the enhancer address established for 7SK (PMID 26878240) belongs to the RNA and its transfer to MEPCE is an inference, not a measurement.
epigenetics not applicablefalse
not applicable rejected becauseThe particle MEPCE stabilises occupies enhancers genome-wide and repositions nucleosomes via BAF, so a chromatin measurement for MEPCE exists to be made.
non histone substrate7SK snRNA 5' gamma-phosphate. This is the field spelling the corpus already uses on 16 entries for exactly this situation.
uniprotQ7L2J0
shown for fusionNO - nobody has looked in any KMT2A fusion.
metric test caseThis entry is over-credited by _EPI_KEY (any key containing 'methylat' scores it covered, on an RNA substrate) and under-credited by _EPI_MARK (its real chromatin relevance names no histone residue). Both errors on one entry, on different lines of the same report.
marks named
[]
measurements
[
  {
    "quantity": "node files naming MEPCE with no entry-level epigenetics record",
    "value": 9,
    "units": "node files",
    "subject": "word-boundary naming count for MEPCE across the corpus, against gene-mepce carrying no epigenetics field",
    "universe": "3,032 node files at commit ac2aeab, via scripts/beat-inventory.py 2026-09-24",
    "null": null,
    "note": "the least-discussed of the three 7SK subunits in the real-gap floor; LARP7 is named 22 times, HEXIM2 10"
  }
]
Still open
  1. Is there a cryo-EM or crystal structure of MEPCE within the assembled 7SK snRNP that tests the catalytic-centre occlusion model? No post-2009 structural search was run.
  2. Does MEPCE have RNA substrates other than 7SK? The paralogue BCDIN3D caps pre-miR-145, and no search was run for MEPCE substrate breadth; a second substrate would change what a 'methyltransferase' field on this entry means.
  3. Would an author writing methyltransferase_activity on gene-mepce be scored as epigenetics-covered by the current reporter? On reading the regex, yes - and testing that on a live run is the cheapest demonstration available of the key-versus-value defect.
  4. Is MEPCE or 7SK capping altered in KMT2A-rearranged blasts? Nothing has been measured.
Every source behind this entry (1)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MEPCE (the 7SK methylphosphate capping enzyme) caps 7SK as a LARP7-free enzyme and probably co-transcriptionally, before the RNA's sequestration into 7SK snRNP; on interacting with LARP7 within the particle MEPCE loses capping activity, probably through occlusion of its catalytic centre; MEPCE then displays a capping-independent function promoting the LARP7-7SK interaction, and MEPCE and LARP7 act cooperatively to stabilise 7SK and maintain the integrity of 7SK snRNP.

    A capping-independent function of MePCE in stabilizing 7SK snRNA and facilitating the assembly of 7SK snRNP · Xue Y, Yang Z, Chen R, Zhou Q · Nucleic Acids Research 38(2):360-9 2009
    PMID 19906723DOI
    What this evidence cannot show

    Abstract-level read, HeLa cells, 2009. Both mechanistic steps are hedged in the source ('probably in a co-transcriptional manner', 'probably due to the occlusion of its catalytic center') and the hedges are preserved above. The paper is cited for MEPCE's substrate and phase behaviour; it is not a chromatin paper and expected_absent_terms says so, because the absence of histone vocabulary in it is part of what this entry asserts.

    Evidence that argues against this (2)
    • The occlusion model is the paper's own inference from loss of activity, not a structure, and later cryo-EM work on the 7SK snRNP could support or overturn it. No search for post-2009 structural work on MEPCE within the particle was run in this beat - the PubMed query used was 'MEPCE 7SK snRNA capping enzyme stability' (4 records) - so this is an unchecked field rather than a clean one, and the entry's step 2 should be read as the best available account rather than a settled one.
    • No paper was searched for on whether MEPCE has any non-7SK substrate. BCDIN3-family enzymes methylate other small RNAs, and BCDIN3D (a paralogue) caps pre-miR-145, so the assumption that MEPCE's only relevant activity is on 7SK is an inference from the name and from one paper's focus. If MEPCE has a second RNA substrate the 'not a histone mark, therefore bounded absence on chromatin' framing still holds, but the entry's implied completeness does not.
Connected to