F13B encodes the noncatalytic B subunit of plasma factor XIII. The evidence separates secretion and assembly of the B subunit from transglutaminase chemistry performed by F13A1. B-containing A2B2 is the circulating proenzyme assembly; the activated A-chain complex is a different state.
Thomas et al., PMID:26247044 provides the most directly inspected B-specific experiments. Wild-type and mutant pEZ-MO1-FXIIIB constructs were expressed in HEK293T cells. The investigators collected cleared extracellular medium and lysates, measured B antigen, and added recombinant A2 to secreted material to measure reconstituted A2B2 antigen relative to B antigen. Methods “Cell culture, transfection, and protein expression and quantification” and “Expression and reconstitution study,” Results of the same names and Figures 1/3 captions were read. These are secretion and assembly assays, not an assay of B catalysis. The confocal ER/Golgi experiments concern biosynthetic transit and mutant retention. Figure captions and their reported quantitative results were read; pixels and supplementary images were not inspected. The genuine normal cache contains complete main text; a new XML request used an unacceptable response type and failed, without affecting that available original.
PMID:2334637(https://doi.org/10.1111/j.1365-2141.1990.tb02585.x) is available here only as its original clinical abstract. It reports familial human B deficiency, reduced plasma A and a short disappearance half-life of infused A, while platelet A remained normal. This is positive carrier/stability evidence with uninspected full Methods. It is not an experiment showing B makes fibrin cross-links.
The full PMID:27363989 article and supplement were recovered in an official Bonn thesis. Its reproduced article begins with “Coagulation Factor XIIIA Subunit Missense Mutations Affect Structure and Function at the Various Steps of Factor XIII Action.” The inspected Methods express sequence-verified human F13A1 constructs in COS-1 cells. Results examine A-chain activity, alpha-2-antiplasmin incorporation, fibrin cross-linking and clot properties. The supplementary alpha-2-antiplasmin assay uses commercial A2B2 (Fibrogammin) as a positive reference preparation. It does not isolate B-chain chemistry. The main Introduction describes B as the protective carrier and describes its release during activation. Main text, selected captions and supplementary assay Methods were read; no figure pixels or unrelated thesis chapters were inspected.
The current ComplexPortal CPX-6231 explicitly identifies two P00488 and two P05160 subunits, with F13A1 designated as enzyme and the overall complex evidence labeled curator-inferred. Its “XIIIa” name/function wording must not erase the inactive-versus-activated assembly distinction. Whole-complex positive controls may legitimately support component-level process participation. Absence of an isolated B perturbation does not establish an evidence-code error. The imported F13B clot-formation IDA is retained with its explicit complex-level scope, and the unsupported wrong-evidence-code defect record is removed. The B participation and A2B2 association remain supported independently. No source code, term or normal record is rewritten, and no upstream report has been submitted.
The selected original author rendering of PMID:14718574(https://doi.org/10.1074/mcp.M300127-MCP200), Data Sources and Methods and Table II (printed pages 315–316), contains the exact P05160 B-chain entry. Its row is positive in the literature, 2DEMS and LCMS1 sources and negative in LCMS2. The study combines human plasma/serum datasets and literature; this is not a new targeted secretion experiment. The target row does not resolve unique peptides, intact mature-chain boundaries or predicted products. The saved text contains selected original portions, so full-text availability is not claimed. PMID:4405643(https://pubmed.ncbi.nlm.nih.gov/4405643/) has verified bibliographic identity but no complete original body here. Its existing extracellular assertion is retained using independent positive evidence, without claiming a new read of its plasma/platelet purification experiments or placing B inside platelets.
The actual PANTHER PTHR45785 tree places human P05160 at PTN002499773 below the positive coagulation IBD PTN001455694. The target MSA sequence is exactly the displayed 661-residue sequence. No target-path IBD NOT contradicts coagulation. Human P05160 in the IBD evidence is legitimate experimental descendant grounding, not circularity. The broader family name and unrelated GO-slim fields are not evidence that F13B performs complement chemistry.
Current mouse Q07968, MGI:88379 and ENSMUSP00000027615 are donor identifiers, not human assay constructs. Its coagulation IMP points to PMID:18224415(https://doi.org/10.1007/s12185-007-0005-z). The actual original abstract reports A- and B-deficient mice and baculovirus-produced recombinant human B rescue of B-deficient plasma A levels, fibrin cross-linking and amine incorporation. Full Methods, reagent sequence and figures were unavailable; these details are not inferred from the abstract. Other propagated mouse annotations are not substituted for this experimental source. Human B assembly supplies independent target-specific support.
The official Secreted/SL-0243 mapping leads to extracellular region. P05160 carries this location with the two experimental secretion/plasma references. In Reactome R-HSA-140599, native complexes 140608 and 140789 contain human B entity 140605; that entity itself is P05160, residues 21–661, extracellular region. Thrombin is the proteolytic catalyst acting on the A chains. In R-HSA-140847, B is released from the cleaved tetramer, while active output 140849 contains A chains and calcium. This is explicit target location and state evidence, not an inference that B performs the activation reaction. The event’s cited 1985 paper was not independently read.
The official displayed sequence is 661 residues and matches NP_001985.2. Signal peptide 1–20 and chain 21–661 are processing features, not splice products. UniProt has no alternative-products comment, and no such field is added. The 18 predicted XP records and 20 captured Ensembl transcript models remain separate records without inferred secretion, assembly or functional equivalence. The expression experiments do not identify all these models; paper-numbered mutations are not mechanically remapped to current coordinates.
Reference availability flags describe whether a complete original main text was recovered. They do not mean every paragraph, figure pixel or supplement was read, and they are independent of immutable normal-cache metadata. Complete main text was available for 26247044 and 27363989; the other three imported papers and the added mouse-donor paper retain their actual access limits.
The one core describes noncatalytic B-dependent assembly and maintenance of the plasma A-subunit pool. It uses the already supported fibrin-clot formation process, extracellular region and broad transferase-complex context. Blood coagulation is an ancestor of that process, so it is not repeated in the same core. The complex label applies to precursor assembly, not persistent B membership in the activated A-plus-calcium state. No protein-glutamine transferase activity is assigned to B, even with a contributes_to qualifier.
The decisive B-specific original is PMID:26247044(https://pmc.ncbi.nlm.nih.gov/articles/PMC4521963/): Methods on cell culture/expression and expression/reconstitution; Results on secreted antigen and heterotetramer assembly; Figures 1 and 3 caption text. Those passages were independently read for synthesis. Added recombinant A2 and secreted B are measured by heterotetramer antigen relative to total B; this demonstrates assembly rather than a catalytic assay or a measured delivery route. The clinical and donor-rescue sources remain abstract-only. The earlier 2016 A-subunit original and actual Reactome component objects preserve the A-versus-B and activation-state distinction. No new original acquisition or image-pixel read occurred in this phase.
An actual bounded official term search identified protein carrier activity, GO:0140597, which requires direct binding and delivery to an acceptor or specific location, and protein-containing complex stabilizing activity, GO:0140777, which requires preventing an interaction with a dissociation-promoting factor. The inspected experiments do not establish those specific mechanisms. Those two mechanism-specific terms remain unassigned. The broader structural molecule activity term is evaluated and assigned below; contributes_to_molecular_function remains omitted because B is released before A catalysis. The former term-assignment-only knowledge-gap object is removed; term selection belongs in these notes rather than being relabeled as an experimental biological unknown. This is not a claim of universal ontology absence or of unknown F13B biology. A mechanism-specific term can be assigned when its defining step is supported. No annotation is added merely to fill a validation field.
All imported source objects, actions and product boundaries are preserved. Row 9 retains positive component-level process participation without a source-defect claim. All reference identities and original-access flags are preserved. The 661-residue displayed precursor, its 1–20 signal/21–661 chain, predicted products and historical constructs remain distinct. Ten selected normal dependencies and two machine inputs are carried forward unchanged from the accepted 22-source union; their prior fixed-cut classification is not a new main observation.
The current GO:0050821 protein stabilization definition concerns maintaining protein structure and integrity and preventing degradation or aggregation. Its direct parent is GO:0031647 regulation of protein stability. It is not an ancestor or descendant of either blood coagulation or fibrin-clot formation, so redundancy would not prohibit a justified annotation. It is a biological-process term and would not itself supply the omitted molecular-function slot.
F13B is a plausible performer, not the substrate of the proposed stabilization: its binding/assembly contribution acts on A. However, the directly inspected human reconstitution experiment measures A2B2 antigen relative to B antigen (Methods “Expression and reconstitution study,” corresponding Results and Figure 3 caption). The human deficiency abstract measures disappearance of infused A, and the human-B rescue in mice reports plasma A abundance and cross-linking. The complete Methods of those historical studies remain unavailable. These observations establish the positive maintenance role but do not distinguish clearance of intact A from protection of A against degradation or aggregation. A NEW assertion of GO:0050821 is therefore deferred on this evidence, without asserting that the function is absent.
The comparator check does not justify automatic gap filling. At this observation, exact-term/descendant QuickGO queries returned no GO:0050821 annotation for human VWF/P04275, the circulating factor-VIII protective partner, or IGFALS/P35858, the IGF-complex subunit associated with prolonged circulating half-life. AHSP/Q9NZD4 is a useful positive contrast: it carries the term and prevents alpha-hemoglobin aggregation/precipitation. The first two are closer plasma-persistence comparators; the third has a directly different anti-aggregation mechanism. Annotation absence does not prove a prohibition, but it prevents inferring that all protective circulating partners routinely receive the term. The pinned repository GO-CAM index and cached models contain no P05160/F13B entry; that finite search supplies neither a contrary modeled role nor evidence for a new assertion.
No publication quotations are added under the explicit campaign instruction. Bare reference support is paired with public source locators. Availability flags continue to report complete original-main access, independently of the cache's extraction metadata. The new precursor-complex question concerns ontology representation; it does not change the observed A/B activation-state distinction or assert B catalysis.
Current GO:0005198 structural molecule activity, independently checked through QuickGO and OLS, describes contribution to the structural integrity of a complex. Its broad scope does not require the special acceptor-delivery mechanism of GO:0140597 or dissociation-factor antagonism of GO:0140777. Neither more specific mechanism is asserted here.
The independently reread PMID:26247044(https://pmc.ncbi.nlm.nih.gov/articles/PMC4521963/) Methods express human wild-type and variant B in HEK293T cells, add 0.75 micrograms recombinant A2 to 200 microliters secreted medium, and quantify reconstituted A2B2 antigen relative to total B antigen. The reporter control showed no significant transfection differences. The corresponding Results and Figure 3 caption report reduced reconstitution for multiple B mutations, with Leu116Phe and Val217Ile comparable to wild type. The B-normalized complex-assembly phenotype provides target-specific evidence for a broad noncatalytic structural contribution. It is more informative than an unidentified co-association, while still being indirect evidence of integrity rather than a measured dissociation lifetime or architecture. The NEW row therefore uses IMP for the mutant comparison and explicitly limits the claim to precursor assembly. No figure pixels, complete-paper reread or new structural analysis are claimed.
The same-role comparator search found no exact GO:0005198 annotation for F13B, VWF, IGFALS or IGFBP3. Descendant queries returned four VWF assertions to GO:0005201 extracellular matrix structural constituent, and none for the other three. VWF's ECM role is distinct from its FVIII-carrier role, so these positives do not establish a carrier-wide convention. The absences do not veto a target-supported function either. Official protein identities were verified; this comparison informs the boundary but is not experimental evidence for F13B. No existing F13B molecular-function assertion is an ancestor or descendant of the proposed term.
The assignment does not alter the actual Reactome activation-state distinction: B leaves before the A-plus-calcium active complex performs cross-linking. It names B's own broad structural contribution to A2B2, not a contribution to A's catalytic molecular function. The previously deferred GO:0050821 assertion remains deferred because assembly and plasma abundance alone do not identify protection against degradation or aggregation. All 11 imported annotations and all 11 complete reference objects remain unchanged; one new structural-function assertion is appended, with no new publication or quotation.