K9IIP0 Research Notes

Key findings

UniProt names this protein Tumor necrosis factor-inducible gene 6 protein [file:DESRO/K9IIP0/K9IIP0-uniprot.txt "RecName: Full=Tumor necrosis factor-inducible gene 6 protein"].
Deep research identifies K9IIP0 as TNFAIP6/TSG-6 with Link and CUB domains [file:DESRO/K9IIP0/K9IIP0-deep-research-falcon.md "UniProt accession K9IIP0 corresponds to tumor necrosis factor alpha-induced protein 6 (TNFAIP6/TSG-6) from Desmodus rotundus (vampire bat), with hallmark Link and CUB domains and hyaluronan-binding/link module and CUB superfamily annotations."].
Deep research notes TSG-6 is secreted and inflammation-inducible [file:DESRO/K9IIP0/K9IIP0-deep-research-falcon.md "TNFAIP6/TSG-6 is a secreted, inflammation-inducible glycoprotein"]

2026-07-31 compliance review

The UniProt CAUTION was being over-read. The previous version used
[file:DESRO/K9IIP0/K9IIP0-uniprot.txt "CAUTION: Lacks conserved residue(s)
required for the propagation of"] to justify UNDECIDED on hyaluronan binding and
on secretion, and REMOVE on the anti-inflammatory term. But the caution is
scoped to PROSITE-ProRule:PRU00059, and PRU00059 is the CUB rule (it maps
to PROSITE PS01180; confirmed at prosite.expasy.org/unirule/PRU00059). The Link
module features in the same record cite a different rule, PRU00323. So the
caution constrains CUB-derived transfer only — not the Link module, not
secretion, not hyaluronan binding. Same pattern as K9IWX5, where the caution was
scoped to the ShKT rule.

Cross-checked every call against human TNFAIP6 (P98066) in QuickGO:

Term	Human evidence	Old call	New call
`GO:0005540` HA binding	IDA ×2 (26468290, 26823460)	UNDECIDED	ACCEPT
`GO:0005615` extracellular space	`GO:0005576` IDA (1730767)	UNDECIDED	ACCEPT
`GO:0050728` neg. reg. inflammatory response	IDA (21569482) + IBA	REMOVE	ACCEPT
`GO:0016787` hydrolase activity	`GO:0106435` IDA ×2	REMOVE	MODIFY → `GO:0106435`
`GO:0007155` cell adhesion	IEA only, no experiment in any species	OVER_ANNOTATED	OVER_ANNOTATED (kept)

The hydrolase reversal is the substantive one. The old reason was "No
evidence of hydrolase activity; annotation is keyword-based and unreliable."
That is factually wrong: TSG-6 catalyses covalent transfer of
inter-alpha-inhibitor heavy chains onto hyaluronan through an ester
intermediate. The UniProt record for this protein says so —
[file:DESRO/K9IIP0/K9IIP0-uniprot.txt "is required for transesterification of
the HC to hyaluronan."] — the deep research independently lists
[file:DESRO/K9IIP0/K9IIP0-deep-research-falcon.md "enzymatic transesterase
activity transferring I"]αI heavy chains among TSG-6's activities, and human
TNFAIP6 carries GO:0106435 carboxylesterase activity from two direct assays
(PMID:16873769, PMID:20463016). GO:0106435 is a descendant of GO:0016787
(verified via OLS ancestors), so the right move is MODIFY, not REMOVE.

Added GO:0030212 hyaluronan metabolic process as NEW (ISS) — the BP
counterpart of that catalytic step, IDA-supported in human, and absent from the
DESRO GOA record.

The Vampirome study (PMID:23411029, cached) is the source of the EMBL record
(JAA46157.1) and confirms the protein in the gland proteome PMID:23411029 — 101 ions, the fourth most abundant family (semaphorin 855 ions >
plasminogen activator 597 > lipocalin 163 > TSG-6 101), corrected from an
earlier "third" in this file and in the review. Worth following up: TSG-6
PMID:23411029-α-inhibitor, while the best-characterised vampire bat salivary protein
is a plasminogen activator. Whether those two are functionally coupled is
recorded as a knowledge gap.

Review follow-up (2026-08-01)

Addressing the PR #2362 review:

core_functions now carries a second entry with molecular_function: GO:0106435 (carboxylesterase activity), directly_involved_in: GO:0030212,
locations: GO:0005615. Previously the file argued in existing_annotations
that the catalytic function was real enough to rescue the hydrolase
annotation, but never listed it as a core function. GO:0030212 moved to
that second entry, where it is the BP counterpart of the catalytic MF.
Promoted the orthologue evidence out of bare in-prose PMIDs into
references: with verbatim findings — PMID:20463016 [Sanggaard et al. "In
concert with HC2, TSG-6 have a unique catalytic activity transferring HCs
from bikunin proteins to hyaluronan (HA)."], PMID:26468290 [Briggs et al.
"TSG-6 was shown to play a direct role in the transfer of HCs from IαI onto
HA via the formation of covalent intermediates"], PMID:16873769 [Forteza et
al. "potentiating the antiplasmin activity of this serine protease
inhibitor"] — and cited them in supported_by for the GO:0016787 MODIFY
and the GO:0030212 NEW.
Re-anchored the GO:0030212 ISS annotation from the deep-research falcon
file to PMID:20463016, since the stated justification is orthologue transfer.
Replaced the domain-conservation quote (which spoke to Link/CUB retention,
not hyaluronan metabolism) with quotes on covalent HC transfer onto HA.
Corrected the abundance rank: TSG-6 is fourth, not third, in the PS gland
proteome (semaphorin 855, plasminogen activator 597, lipocalin 163, TSG-6
101). Fixed in the PMID:23411029 finding, in the knowledge-gap
significance, and above in this file.

Also from the non-blocking suggestions: the transesterification-vs-hydrolysis
mismatch of GO:0106435 is now recorded in the MODIFY reason and as a
suggested_questions entry rather than only in the PR description, and the
GO:0005615 reason no longer treats gland-tissue LC-MS/MS as localisation
evidence. Left unchanged: the GO_REF findings quote the UniProt flat file
rather than the GO_REF document — real provenance nit, but rewriting those
findings would move quotes away from the reference they document.

just validate DESRO K9IIP0 → ✓ Valid.