C1GALT1C1 (COSMC) — Function-Assignment Hypothesis Report OpenScientist openscientist-autonomous 7 citations 2 artifacts 2026-09-17T05:41:38.402808 citations file

OpenScientist final report (text/html)

OpenScientist final report (application/pdf)

C1GALT1C1 (COSMC) — Function-Assignment Hypothesis Report

Gene: C1GALT1C1 (COSMC) · UniProt: Q96EU7 · Organism: Homo sapiens (NCBITaxon:9606)
Term under review: GO:0016263 — N-acetylgalactosaminide beta-1,3-galactosyltransferase activity (MF)
Existing annotation: IMP, PMID:37216524, negated = true
Seed hypothesis: C1GALT1C1 does not have N-acetylgalactosaminide beta-1,3-galactosyltransferase activity (GO:0016263).

Executive Judgment

Verdict: SUPPORTED (high confidence). C1GALT1C1 encodes Cosmc, a private ER molecular
chaperone that is required for the folding/stability of the core 1 β3-galactosyltransferase
(C1GALT1 / T-synthase, UniProt Q9NS00). The catalytic activity GO:0016263 (synthesis of
Galβ1-3GalNAc-α-Ser/Thr, the T antigen) belongs to C1GALT1, the client enzyme — not to
C1GALT1C1. Multiple lines of primary evidence, and an explicit UniProt CAUTION, show that
C1GALT1C1 has no transferase activity. The negated annotation is therefore justified and should
be retained.

Key caveat on evidence code: the annotation uses IMP with negation. The mutant phenotype in the
reference (PMID:37216524) shows loss of T-synthase (C1GALT1) activity when Cosmc is defective. That
result demonstrates Cosmc's chaperone requirement for the reaction; it does not show Cosmc itself
catalyzes it. IMP is defensible for a NOT-annotation here, but the positive function of this gene is
chaperone activity, and that is what should carry the informative MF term.

Evidence Matrix

Citation	Evidence type	Stance	Claim tested	Key finding	Context	Confidence / limitations
UniProt Q96EU7 (CAUTION)	Database/curation	Supports	Does C1GALT1C1 have GT activity?	"Was originally assigned to be a glycosyltransferase. However, it was later shown that it has no transferase activity and rather acts as a chaperone." KW=Chaperone (no Transferase).	Human, reviewed entry	High; database-level, but cites primary work
Ju & Cummings 2002, PMID:12464682	Direct assay / mutant complementation	Supports	Is the transferase activity intrinsic to Cosmc?	Cosmc restores C1β3Gal-T activity in Cosmc-deficient Jurkat & insect cells; C1GALT1 gene/mRNA is normal in those cells; Cosmc "acts as a specific molecular chaperone in assisting the folding/stability of C1β3Gal-T." The transferase is a separate gene product.	Human Jurkat, Hi-5 insect cells	High; foundational functional study
Aryal, Ju & Cummings 2010, PMID:19923218	Direct in vitro assay	Supports	Does Cosmc perform folding vs catalysis?	Cosmc directly refolds denatured T-synthase in vitro (ATP-independent, substrate-specific); "first ER chaperone required for folding of a glycosyltransferase." Tn-syndrome Cosmc mutant has reduced chaperone function.	Recombinant proteins in vitro	High; isolates chaperone activity
Aryal, Ju & Cummings 2014, PMID:24616093	Interaction / mechanism	Supports	Mechanism of Cosmc–client recognition	Cosmc recognizes a specific stem-region motif (CBRT) in T-synthase; grafting CBRT onto β4GalT1 confers Cosmc binding — a chaperone recognition system, not catalysis.	Human proteins, in vitro	High
Erger et al. 2023, PMID:37216524 (the IMP reference)	Mutant phenotype (germline)	Supports	What does a germline C1GALT1C1 variant cause?	Titled "Germline C1GALT1C1 mutation causes a multisystem chaperonopathy"; A20D-Cosmc reduces C1GALT1 (T-synthase) activity and impairs O-glycosylation → aHUS-like disease. Loss of chaperone function, not loss of intrinsic transferase activity.	Human patients (2 half-brothers)	High; defines role as chaperone
Crew et al. 2008, PMID:18537974	Mutant phenotype (human)	Supports/qualifies	Consequence of inactivating C1GALT1C1 variants	Inactivating variants (E152K, S193P, M1I) produce Tn phenotype via T-synthase deficiency — i.e., Cosmc loss disables the transferase, consistent with a chaperone role.	Human erythroblasts	High
This report — pairwise alignment	Computational (evolutionary)	Qualifies	Why was GT activity ever attributed?	COSMC (318 aa) vs C1GALT1 (363 aa) ≈ 33% identity; both in GT31 (β3-Gal-T subfamily). Explains original name "C1Gal-T2" and the GT misannotation (PubMed:12361956).	Sequences Q96EU7 / Q9NS00	Medium; coarse NW alignment, but consistent with UniProt family assignment
QuickGO API (this report)	Database provenance	Supports	Where does GO:0016263 actually sit?	C1GALT1C1/Q96EU7: single GO:0016263 = NOT\|enables, IMP, PMID:37216524 (the annotation under review). C1GALT1/Q9NS00: GO:0016263 = positive enables via IDA (PMID:11677243) + IBA + IEA.	EBI QuickGO, human	High; confirms term is experimentally grounded on the paralog, negated on the chaperone
Ju et al. 2002, PMID:11677243	Direct assay (IDA source)	Supports	Which gene product is the enzyme?	"Cloning and expression of human core 1 β1,3-galactosyltransferase" — establishes C1GALT1 as the GO:0016263 catalyst.	Human, recombinant enzyme	High

GO Curation Implications (leads — require curator verification)

Retain the negated GO:0016263 (MF) on C1GALT1C1. The evidence robustly supports that this gene
product lacks N-acetylgalactosaminide β-1,3-galactosyltransferase activity. The negation prevents
paralog carry-over of the T-synthase activity onto its chaperone.
Positive MF that IS supported: chaperone activity — candidate terms GO:0051082 (unfolded
protein binding) and/or GO:0044183 / GO:0140662-type chaperone terms; most specifically the
role assisting C1GALT1 folding. Direct evidence: PMID:19923218, PMID:12464682. Avoid "protein
binding" as the final term — "unfolded protein binding / chaperone" is more informative.
Supported BP: protein folding (GO:0006457) and O-glycan / T-antigen biosynthesis via enabling
C1GALT1 (e.g., GO:0016266 O-glycan processing) as an involved-in / contributes-to relationship,
not as a direct catalytic step.
Supported CC: endoplasmic reticulum membrane; single-pass type II membrane protein (GO:0005789).
Evidence-code note (lead): consider whether the NOT-GO:0016263 is better supported as IC/ND
or kept as IMP; the mutant data are about the client's activity, so IMP for a NOT-MF is
acceptable but worth a curator note.

GO Decision Table (lead — curator to verify)

GO term	Aspect	Gene	Current / proposed	Recommended action	Evidence
GO:0016263 N-acetylgalactosaminide β-1,3-galactosyltransferase activity	MF	C1GALT1C1	NOT\|enables, IMP, PMID:37216524	RETAIN negation	No intrinsic transferase activity; UniProt CAUTION; PMID:12464682/19923218
GO:0016263 (same)	MF	C1GALT1 (paralog Q9NS00)	enables, IDA PMID:11677243 (+IBA/IEA)	Keep (belongs here) — reference point	Direct assay of the actual enzyme
GO:0051082 unfolded protein binding	MF	C1GALT1C1	(candidate to add/confirm)	ADD / confirm as informative positive MF	PMID:19923218, PMID:12464682, PMID:24616093
GO:0006457 protein folding	BP	C1GALT1C1	(candidate)	ADD/confirm — direct chaperone role	PMID:19923218
GO:0016266 O-glycan processing (or 'contributes to' C1GALT1 activity)	BP	C1GALT1C1	(candidate)	ADD as involved-in/contributes-to, not direct catalysis	PMID:37216524, PMID:24643043
GO:0005789 endoplasmic reticulum membrane	CC	C1GALT1C1	(candidate)	Confirm; single-pass type II membrane	UniProt Q96EU7

Bottom line: retain the negated GO:0016263 on C1GALT1C1; the gene's informative positive molecular function is chaperone / unfolded-protein binding, not any transferase term.

Mechanistic Scope

Immediate molecular function of C1GALT1C1: ATP-independent, client-specific molecular chaperone
in the ER that binds a stem-region motif (CBRT) of C1GALT1 and promotes its folding/stability,
preventing proteasomal degradation.
Not the gene product's activity: the galactosyl transfer reaction itself (Galβ1-3GalNAc-Ser/Thr
synthesis) — that is enzymatically executed by C1GALT1.
Downstream / indirect (not to be conflated with MF): T-antigen formation, extended O-glycan
biosynthesis, Tn/STn antigen exposure on Cosmc loss, and disease phenotypes (Tn polyagglutination
syndrome, aHUS-like chaperonopathy). These arise from loss of chaperone function, not from loss of
an intrinsic transferase.

Conflicts and Alternatives

Historical database carry-over (resolved): the original 2002 report (PubMed:12361956) named the
protein "Core 1 β1,3-galactosyltransferase 2 (C1GalT2)" based on GT31 paralogy (~33% identity to
C1GALT1). This is the source of any residual GT annotation; it was explicitly refuted (UniProt CAUTION;
PMID:12464682). No credible primary evidence supports intrinsic transferase activity.
Paralog confusion (the main risk): C1GALT1 (Q9NS00) is the true GO:0016263 enzyme. Annotation
pipelines keyed on GT31 family membership could re-propagate the term; the explicit negation guards
against this.
No organism-specific or isoform-specific evidence contradicts the chaperone-only interpretation;
the chaperone role is conserved across vertebrates (client-specificity data, PMID:24616093).

Knowledge Gaps

Formal enzymatic exclusion assay. Checked: no primary paper reports a direct in vitro assay of
purified Cosmc acting as a galactosyltransferase; activity has only ever been observed via C1GALT1.
Matters because the negation rests on absence of evidence + mechanistic understanding. Resolved by
an in vitro UDP-Gal → GalNAcα-O-phenyl assay with purified Cosmc alone (expected: no activity).
Structural confirmation of a degenerate catalytic site. Checked: UniProt lists GT31 similarity
but no experimental structure with catalytic-residue analysis surfaced in searches. A structure/model
confirming loss of the catalytic/acceptor-binding residues would strengthen the mechanistic basis.
Evidence-code appropriateness for the NOT annotation. Whether IMP vs IC/ND best fits a negative
MF; a curator judgment, not an experimental gap.

Discriminating Tests

Direct transferase assay of purified Cosmc (UDP-[3H]Gal donor, GalNAcα-O-phenyl acceptor) with
C1GALT1 as positive control — cleanly separates "no intrinsic activity" from "chaperone only."
Reciprocal reconstitution: active T-synthase produced only when Cosmc is co-expressed in
Cosmc-null (insect) cells — already shown (PMID:12464682); a targeted re-run isolates the chaperone
dependency.
Structure/AlphaFold catalytic-residue mapping: compare Cosmc vs C1GALT1 active-site and
donor-binding residues to show catalytic degeneration.

Curation Leads (verify before applying)

Action: Keep negated GO:0016263 (MF) on C1GALT1C1. Do not add any positive galactosyltransferase term.
Add positive MF: unfolded protein binding / chaperone activity (e.g., GO:0051082) — refs PMID:19923218, PMID:12464682.
Add/confirm BP: protein folding (GO:0006457); contributes-to O-glycan biosynthesis (via C1GALT1).
Add/confirm CC: endoplasmic reticulum membrane (GO:0005789), integral/single-pass type II.
Candidate references w/ exact snippets to verify:
PMID:12464682 — "Cosmc acts as a specific molecular chaperone in assisting the folding/stability of C1β3Gal-T."
PMID:19923218 — "Cosmc represents the first ER chaperone identified to be required for folding of a glycosyltransferase."
PMID:37216524 — title "Germline C1GALT1C1 mutation causes a multisystem chaperonopathy."
UniProt Q96EU7 CAUTION — "…it has no transferase activity and rather acts as a chaperone."
Suggested curator question: Is IMP the preferred evidence code for this NOT-MF, or should IC/ND be used given the phenotype reflects the client enzyme's activity?
Suggested experiment: direct in vitro transferase assay of purified Cosmc (expected negative).

Provenance

Literature via PubMed (PMIDs above); entry data via UniProt Q96EU7 (.txt) and Q9NS00 (.fasta).
Computed pairwise global alignment COSMC vs C1GALT1: ~33% identity (318 vs 363 aa), GT31 paralogs
— executed in this run (Needleman–Wunsch over fetched UniProt sequences).