Chimeric mRNA Trans-Fusions in Immunity

Bottom line: a chimeric trans-fusion transcript joins exons from two separate genes,
often on different chromosomes, into one mRNA by RNA trans-splicing with no DNA
rearrangement, and its translation can yield a hybrid protein with its own function. A 2026
Nature study (PMID:42686912) showed that inflammation in mouse macrophages produces a
GSDMD:TMEM106A chimera whose TMEM106A part is read out of frame, and that the chimeric protein
binds canonical GSDMD N-termini to speed pore formation and IL-1β release. We reviewed the two
human parent genes, GSDMD (77 annotations) and TMEM106A (10), both now COMPLETE, and recorded
the chimera in each review's knowledge_gaps and questions rather than as an annotation of
either parent. We did this because GO assumes one gene gives one set of products, and a
chimera breaks that: its function belongs to neither parent, the TMEM106A contribution is not
the canonical protein, and no human chimera has yet been shown. No chimera-level annotation
exists and none is proposed; what remains is the set of open questions at the end of this
page.

Why this matters for gene-function curation

GO annotation, and the reviews in this repository, are organized around one gene → one
(set of) gene product(s). Chimeric trans-fusions break that assumption: a functional
protein can arise from two loci at once, and — as in the flagship case below — the
contribution of one parent can come from an alternative (out-of-frame) reading of its
mRNA rather than its canonical protein. This creates three concrete curation problems:

1. Attribution. A chimera's function is not an annotation of either parent gene. It
   should not be added to existing_annotations or core_functions of GSDMD or TMEM106A;
   it belongs in knowledge_gaps / notes as adjacent biology (this is how the two reviews
   here handle it).
2. The "wrong-frame" trap. The TMEM106A portion of GSDMD:TMEM106A is a cryptic peptide,
   so naïvely transferring canonical TMEM106A function to the chimera (or vice versa) would
   be wrong.
3. Ortholog scope. The effector was characterized in mouse; whether a human
   orthologous chimera exists and is functional is an open question, so human GSDMD/TMEM106A
   reviews must flag it as unresolved rather than assert it.

Flagship example: GSDMD:TMEM106A (Gsdmd-Tmem106a)

A 2026 study established that regulated transcript fusion produces functional proteins during
inflammation, with a trans-spliced GSDMD:TMEM106A chimera as the worked example
[PMID:42686912].

• What it is. "we identify a protein-coding chimeric mRNA representing a fusion between
  the pore-forming protein gasdermin D (GSDMD) ... and a C-terminal domain translated out of
  frame from Tmem106a (Gsdmd-Tmem106a) in mice" [PMID:42686912]. The GSDMD contribution is
  its pore-forming portion; the TMEM106A contribution is an out-of-frame peptide, i.e.
  not the canonical TMEM106A protein.
• How it forms — literally "trans". The two parent genes sit at distinct, non-adjacent
  loci and the fusion is made at the RNA level by trans-splicing, not by a DNA
  translocation or a cis read-through: "Chromatin conformation capture studies reveal that
  inflammation induces interchromosomal DNA interactions, positioning parent genes
  proximally to facilitate the formation of chimeric mRNA" [PMID:42686912]. The
  characterized chimera is the mouse Gsdmd/Tmem106a pair; we have not verified the
  mouse chromosome assignments, and the abstract does not state the splice-site chemistry,
  so no more specific mechanism is asserted here.
• When/where. Inflammasome priming upregulates the chimera in myeloid cells; the protein
  localizes to the plasma membrane [PMID:42686912].
• Function. After inflammasome activation, "GSDMD-TMEM106A directly interacts with
  canonical GSDMD N termini to accelerate and enhance pore formation and IL-1β release"
  [PMID:42686912] — i.e. it is a membrane cofactor that speeds up canonical GSDMD-NT pore
  assembly and pyroptosis.
• In vivo. "GSDMD-TMEM106A balances host defence and immunopathology in vivo: its loss
  protects against lethal sepsis but compromises antibacterial defence, whereas
  overexpression enhances host protection while increasing sepsis lethality" [PMID:42686912].
• Scale. The study runs a discovery pipeline — "long-read direct RNA sequencing with
  non-targeted and targeted validation to identify chimeric transcripts in macrophages"
  [PMID:42686912] — so GSDMD:TMEM106A is the worked example out of a larger catalogue. The
  cached record for PMID:42686912 is abstract-only (full_text_available: false) and the
  abstract gives no catalogue size, so no figure is quoted here. (An earlier draft of this
  page asserted ">30,000 chimeric mRNAs"; that number is not in any source we hold and has
  been removed.)

Distinguishing chimeric trans-fusions from look-alikes

Note on the examples. Only the GSDMD:TMEM106A row is sourced here. An earlier draft
listed CLEC12A-MIR223HG in the trans-spliced row; PR review flagged that the two loci
are neighbours on 12p13.31, which would make a cis read-through the more likely
reading. Since the entry carried no citation either way, it has been removed rather than
reclassified on an unsourced argument. The other illustrative pairs were likewise uncited
and have been dropped; the table now states the criteria that distinguish the three
phenomena, which is what it is for. Re-add examples only with a citation that establishes
the mechanism, not just the fusion.

The GSDMD:TMEM106A case is notable for being a physiologically functional immune effector,
whereas many catalogued chimeras are cancer-associated or of unproven function.

Genes reviewed here

• GSDMD — pyroptosis executioner; the pore-forming parent. See its review's
  knowledge_gaps for the cofactor / human-chimera question.
• TMEM106A — plasma-membrane macrophage-activation regulator; the parent whose locus
  contributes the out-of-frame peptide. Its review flags both its own dark molecular function
  and the human-chimera question.

Open questions

• Does a human GSDMD:TMEM106A chimera exist and function as in mouse?
• What sequence/chromatin features specify which transcript pairs are trans-spliced during
  inflammation, and how is the process regulated?
• How many of the catalogued chimeras are translated and functional, and by what
  criteria should any be curated as distinct gene products for GO?
• Do other pyroptosis/inflammasome components participate in functional chimeras?

References

• PMID:42686912 — Venezia O, Kane H, et al. (senior author R. Jackson). Functional
  chimeric mRNAs encode proteins in mammalian immunity. Nature, 2 Sep 2026.
  DOI: 10.1038/s41586-026-10982-x. (Primary source; PubMed-verified.)

Slides

• Slides (Marp source: CHIMERIC_MRNA_IMMUNITY-slides.md) — AI generated