Reviewed as part of the contested-functions batch. Headline finding: GO got this one right.
The contested claim (histone tyrosine sulfotransferase activity) never entered GOA, and nothing
in the existing annotation set needs to change because of the dispute.
SULT1B1 is one of the ~13 human cytosolic (SULT-family) sulfotransferases. It transfers the
sulfonate group of PAPS to small phenolic acceptors and to iodothyronines.
GO:0006576 biogenic amine metabolic process (TAS) annotation is kept rather than acceptedCytoplasm with ECO:0000269 from both foundingcytosol TAS annotations are the more informative form.GO:0006068 ethanol catabolic process (IDA, PMID:23207770, assigned by CAFA). The cited paper's own
abstract positively excludes SULT1B1 from the set of ethanol-sulfating enzymes:
PMID:23207770
SULT1B1 was one of the eleven assayed and was a negative. I have marked this
MARK_AS_OVER_ANNOTATED rather than REMOVE (project rule: do not remove an experimental
annotation whose full text I have not read), but I regard it as a probable mis-transfer of the
paper's panel result. The sibling annotations from the same paper (GO:0004062, GO:0051923,
GO:0050427) are fine — SULT1B1 genuinely was a purified, PAPS-using enzyme in that panel.
GO:0030855 epithelial cell differentiation (IEP, PMID:21492153) comes from a 2-D gel proteomic
comparison of proliferating vs differentiated Caco-2 cells. That is an expression correlation, not a
role in the process; over-annotated.
Youssef et al. reannotated the original raw MS data (PRIDE PXD043754), ran synthetic-sulfopeptide
controls, repeated the enzyme assay, and blotted with an anti-sulfotyrosine antibody.
A gastric-cancer paper attributes the same H3Y99sulf mark to a different enzyme, GAL3ST1:
PMID:41686426
It cites the 2023 SULT1B1 paper as established background and explicitly proposes an enzyme swap:
PMID:41686426
I searched the full text of PMID:41686426 for "Youssef" and "refute": the refutation is not cited.
That paper's H3Y99sulf readouts rest on the anti-H3Y99sulf antibody and immunoblotting — exactly the
reagent and readout the refutation says need independent validation. See the GAL3ST1 notes.
Yes. GO:0008476 protein-tyrosine sulfotransferase activity exists
("3'-phosphoadenosine 5'-phosphosulfate + protein tyrosine = adenosine 3',5'-bisphosphate + protein
tyrosine-O-sulfate"), and GO:0006478 peptidyl-tyrosine sulfation exists on the BP side. A QuickGO
annotation query (goId=GO:0008476, goUsage=exact, taxonId=9606, retrieved 2026-09-17) returns 24 rows,
all of them TPST1 (O60507) or TPST2 (O60704) plus their isoform accessions — the two Golgi
tyrosylprotein sulfotransferases. Neither SULT1B1 nor GAL3ST1 appears.
So the ontology is fully capable of expressing the disputed claim; GO simply has not asserted it for
either enzyme. That is the correct outcome for a contested finding, and it is the finding of this
review.
Core: cytosolic aryl/phenol sulfotransferase (GO:0004062) acting in the cytosol (GO:0005829) on
small phenols and iodothyronines. Non-core: PAPS metabolism, flavonoid metabolism, biogenic amine
metabolism (conflicting evidence). Over-annotated: bare protein binding (5 IPI rows from
high-throughput interactome screens), ethanol catabolic process, epithelial cell differentiation.
Generalized parents (GO:0008146, GO:0006790) modified to the specific terms.
No annotation action was changed on account of the histone-sulfation dispute, because GOA carries
no histone-sulfation annotation for this gene.