Human blood-filtrate analysis identified glycosylated endostatin and related circulating fragments; it does not establish circulation of every intact collagen-XVIII splice product.

A human Knobloch pedigree carried a COL18A1 splice-site defect affecting the short transcript, with ocular abnormalities and occipital encephalocele. The abstract links this to retinal structure and neural-tube closure without resolving a specific skeletal mechanism.

Human aortic NaCl, guanidine and SDS fractions contained the COIA1_HUMAN group with 11, 45 and 6 unique peptides, respectively. These are localization measurements, not a tensile-force assay.

This collagen-family review describes collagen XVIII as a multiplexin/proteoglycan with a carboxy-terminal trimerization module and potentially distinct cleaved fragments; its synthesis is not a new target-specific mechanical experiment.

The human COL18A1 group was detected in feeder-free HUES1-derived matrix at normalized spectral count 5.83, with ND in the six feeder columns. This does not establish absence in all feeder or tissue contexts.

The original review Table 1 lists the collagen-XVIII C-terminal endostatin product with vascular elastic fibres. This is a fragment-level association, not localization of every intact precursor or an elastic-fibre assembly assay.

The original complete-proteome and matrisome worksheets repeat the same COL18A1 protein group in human colon/liver samples at row 110. The accession grouping does not resolve individual splice products.

Endostatin preparations inhibited endothelial proliferation and angiogenesis; the abstract explicitly describes bacterial recombinant material as a nonrefolded suspension. Activity is therefore recorded with preparation scope.

Recombinant human collagen-XVIII NC1 and engineered endostatin oligomers elicited matrix-dependent endothelial movement; monomeric endostatin did not reproduce the motogenic effect. Downstream signaling dependence does not establish collagen kinase or receptor activity.

Purified native circulating human endostatin lacked antiproliferative activity in the tested endothelial-cell assays. This limits generalization across preparations without disproving other fragment effects.

Human amino-terminal collagen-XVIII variants differ in promoter/splicing architecture; the historically named NC1-728 variant contains a frizzled-related cysteine-rich region and has tissue-specific localization. These historical names are distinct from the excised carboxy-terminal NC1 domain.

An isolated 54-residue human collagen-XVIII domain formed a compact trimer after removal of the fusion helper. Full-NC1 sedimentation measurements used mouse proteins; neither experiment measured whole-human-precursor tensile strength.

Added endostatin had positive vascular effects in mouse embryoid-body assays, including increased endothelial proliferation, migration and tube formation, alongside apoptosis and morphological changes. The abstract supports context-dependent rather than uniformly inhibitory activity.

Mouse Col18a1 loss altered retinal vascular development and vitreoretinal attachment. Basement-membrane immunostaining with knockout specificity controls supports a structural anchoring role without defining current human isoform behavior.

Serum from ticlopidine-treated rats altered HUVEC proliferation and apoptosis; anti-endostatin antibody reversed those effects whereas isotype and anti-VEGF controls did not. This supports a contextual drug-associated response, not collagen-mediated drug metabolism.

Mouse Col18a1 loss broadened basement membranes in several tissues and produced strain-dependent hydrocephalus. These observations support structural matrix maintenance without establishing a new catalytic mechanism.