GGCX directly modifies protein glutamate residues and oxidizes vitamin K hydroquinone; both chemical aspects belong to one catalytic function. Human enzyme assays and recent substrate-bound structures support this function at the ER membrane, with the protein-modifying machinery facing the lumen. A lumenal catalytic domain does not make the whole membrane protein a freely soluble lumen resident (PMID1749935, PMID10910912, PMID39880037).
Substrate-recognition structures clarify how propeptides and other recognition sequences position glutamate-rich regions. Proposed bicarbonate/proton-transfer mechanisms remain distinct from direct measurement of every step, and epoxidation-only mutant readouts are not automatically site-specific carboxylation assays (PMID41290650, PMID41888148).
The two official natural sequences have 758 and 701 residues; the shorter product deletes positions 15–71, overlapping the first annotated membrane helix. Its activity and physiological contribution remain unresolved. Initiator removal, experimental tags and engineered fusions are separate issues. The historical two-transcript report concerns 3-prime untranslated sequence rather than demonstrating these two protein forms (PMID9166845).
Mouse Ggcx deletion directly removes precursor-modifying activity, while bleeding and endocrine consequences depend on which substrates and tissues are affected. The accessible evidence leaves five specific neurotransmitter/bone/testosterone contexts unresolved. These are source-specific uncertainties, not claims that the curated experiments were absent or performed on the wrong gene (PMID17327402, PMID25264202, PMID17073445).
The source guide separates primary reading scope, constructs and products. The evidence archive preserves original contexts, both extension conjunctions and the source/sequence/ancestry records.
Carboxylation with osteocalcin as input as part of neurotransmitter suppression is a more specific assertion than enzyme activity alone. The two contextual annotations therefore retain distinct judgments: the vitamin-K input/output conjunction is supported, while the osteocalcin/neurotransmitter conjunction remains unresolved. The inaccessible original PMID:17073445(https://pubmed.ncbi.nlm.nih.gov/17073445/) cannot establish a citation defect merely through its abstract omissions.
Osteocalcin recognition and modification are supported by human GGCX structural and biochemical work (PMID:41290650(https://pubmed.ncbi.nlm.nih.gov/41290650/), PMID:41888148(https://pubmed.ncbi.nlm.nih.gov/41888148/)). That substrate relationship alone does not establish the direction of each neurotransmitter, bone-development or testosterone-biosynthesis endpoint. The earlier PMID:25264202(https://pubmed.ncbi.nlm.nih.gov/25264202/) assessment attributed an observation to publisher highlights that were not retained; the current assessment is limited to the authentic abstract. No endpoint is rejected from an abstract omission.
The 758-residue human enzyme has direct biochemical and structural support. The 701-residue product lacks canonical positions 15–71, overlapping part of the first annotated membrane helix; its activity, membrane topology, expression and physiological contribution remain unresolved. These are biological limits, not established functional classes. The alternative-product records retain both exact sequence identities; constructs and initiator processing remain separate.
The curated GGCX record supports osteocalcin input independently of the unresolved neurotransmitter relationship. Its ACTIVITY REGULATION section also records phosphatidylcholine/cholesterol dependence of optimal carboxylation, citing PMID:39880037 and PMID:39880952. This provides membrane-context evidence without establishing another molecular function.