The sources below distinguish human FXIIa catalysis, surface-dependent activation, processing forms and contextual localization. Original availability is reported separately from the particular experiments inspected. Citations identify the operative sections without adding publication quotations.
PMID:34967109(https://pubmed.ncbi.nlm.nih.gov/34967109/), publisher main, provides independently read human evidence. Methods 2.1–2.7 identify the FXII forms and distinguish three assays. Surface-dependent autoactivation contains human FXII, silica or polyphosphate and zinc, without prekallikrein, followed by an FXIIa-calibrated readout. FXI and prekallikrein activation use human FXIIa and add its inhibitor CTI before the respective downstream chromogenic reporter. Those latter reporters measure FXIa or kallikrein, rather than direct turnover of a chromogenic substrate by FXII. The no-prekallikrein assay supports functional autoactivation without mapping the activation bond or determining the activity of every individual initially single-chain molecule.
Results 3.1–3.4 and their accompanying text/captions distinguish alpha-FXIIa binding to plasma HRG from beta-FXIIa and zymogen under the tested zinc conditions. Reciprocal SPR is positive binding evidence; no universal behavior of all processing forms is inferred. The complete publisher main is accessible separately from the immutable abstract-only normal cache. No new figure pixels or supplementary material were inspected in this lane.
The actual EC 3.4.21.38 entry explicitly identifies factor XIIa and internal protein cleavage. This and the human substrate-activation evidence justify refining broad peptidase activity to serine-type endopeptidase activity. Broad proteolysis is refined to zymogen activation: F12 performs the cleavage that activates FXI or prekallikrein. F12 being cleaved by kallikrein or thrombin is a different role and is not used as evidence that F12 performs those enzymes’ steps. These existing-annotation refinements agree with the direct lane’s specific catalytic and zymogen-activation judgments; no NEW process is added.
PMID:18725990(https://pmc.ncbi.nlm.nih.gov/articles/PMC2518075/) is available as complete original main text. Selected Methods, Results, Discussion and captions were independently read. The misfolded-protein experiments use FXII/prekallikrein omission controls and measure downstream kallikrein. The FXI assay is separate. Tested nonfibrillar aggregates and fibrils behaved differently, so a pathway label containing “amyloid” does not establish activation by every aggregate state. The proposed protective protein-clearance function in Discussion remains a hypothesis. No new pixel or supplement reading is claimed.
The actual PANTHER v19 PTHR24264 tree and MSA place human P00748 at PTN002545599, with an exact 615-residue sequence match. The target descends from all three source-asserted nodes: PTN001208002, PTN002800021 and PTN002545596. Six actual IBD assertions occur on its path, with no path NOT assertion. Self-WITH is legitimate experimental descendant grounding; donor counts are not used as confidence scores.
The coagulation donor MGI:1891012 is mouse F12/Q80YC5. The full author-rendered main of PMID:15351846(https://pubmed.ncbi.nlm.nih.gov/15351846/), original author copy, was recovered. Methods, Results, Discussion and caption text were read; the standalone PDF download failed and no pixels were inspected. Targeted F12 deletion removes the plasma protein and markedly prolongs aPTT. The normal general health and bleeding observations limit a universal physiological-hemostasis claim, but do not contradict participation in the intrinsic contact pathway. The plasma activity measurement is not a purified-protein bond assay. Together with direct human FXI activation, it supports refinement of broad blood coagulation to the intrinsic pathway.
The rough-ER donor MGI:1859281 is mouse Hgfac/Q9R098, a distinct family member. The actual complete publisher main of PMID:12923239(https://pubmed.ncbi.nlm.nih.gov/12923239/), publisher article, was read through Methods, Results, Discussion and captions. Three antibody reagents, antigen absorption controls and N19 immunogold support hepatocyte rough-ER residence. The antibodies do not distinguish active from inactive HGFA, and the authors interpret the intracellular pool as biosynthetic precursor. The verified PAINT ancestry and positive residence inference remain intact. The imported is_active_in assertion is marked over-annotated because the donor localization does not establish activity in that compartment. This is neither a human F12 imaging experiment nor an experimental demonstration of catalytic inactivity. No figure pixels were inspected.
The official GO reference methods for InterPro, phylogenetic inference and combined IEA pipelines were read. Actual target matches and current InterPro2GO entries support IPR014394 proteolysis and IPR001881 calcium-binding EGF domains. The latter intervals are 95–131 and 178–210 in the current target. Calcium binding is retained as a predicted structural facet, not a measured F12 affinity or an independent principal catalytic activity. Zinc conditions in the human activation/HRG assays do not verify calcium binding.
SL-0243 supports extracellular location. All current ARBA00027128 and ARBA00027408 clauses were evaluated against the captured target family, domain and taxonomy fields; neither current rule supplied a complete matching condition. This is an unresolved historical-trigger reconstruction, not proof of a source error or a reason to discard independently supported extracellular and serine-endopeptidase functions.
All 19 imported Reactome roots were traced through 173 fetched participant objects plus six inline terminal chemical leaves. The traversal includes pathway hasEvent, catalyst active units and regulatedBy branches. Initial rate-limit failures are preserved; successful later captures supply the final graph. Thirteen target entities resolve to human P00748, and two separately recovered residue-modification objects verify T328R/K. This is a source-model audit, not an assertion that every cited primary paper was read.
| Source context | Target role and retained boundary |
|---|---|
| 9970672, 9655850, 9909048 | F12 performs prekallikrein activation. Kallikrein performs downstream kininogen cleavage. |
| 158300, 9936430 | Native extracellular alpha-FXIIa is the FXI-activation catalyst, distinct from the receptor-associated substrate complex. |
| 158313, 9909046 | F12 is the substrate of kallikrein. Its extracellular products and receptor-associated location remain supported. |
| 9855735, 9857814, 9655840 | The source assigns an explicit F12 active unit. Independent human chemistry supports enzyme activity, without re-establishing every source processing-site assignment. |
| 158357, 9651456, 9653879 | Native extracellular FXIIa associates with inhibitors/HRG. In 9651456 the disease variant is SERPING1, not F12. |
| 9653249, 9655046 | Actual T328K/R F12 disease precursors undergo thrombin/kallikrein processing. Their locations are retained non-core, without transferring variant processing to ordinary F12. |
| 9855685, 9855705 | Native extracellular F12 participates in surface/polyanion assembly. Assembly and activation are distinct steps. |
| 9936906, 9936909 | Native extracellular target location is retained; the source aggregate label does not erase preparation-dependent activation differences. |
Native source chains are P00748(20–615), alpha-heavy 20–372, light 373–615, and beta-heavy fragment 354–362. Variant processing additionally models 20–328 and 329–615. A T328K/R modification on a processed C-terminal object can record its precursor provenance even though residue 328 is outside that fragment. No sequence repair or new product object is invented.
The native receptor-associated F12 input of R-HSA-158313 supports refinement from plasma membrane to external side of plasma membrane, which includes peripheral external-face association. This attribution rests on the exact source topology. The separately reported gC1qR ligand-blot experiments are not represented as direct intact-cell F12 imaging. The variant-specific membrane row remains a non-core disease context. Explicit review reasons on the three variant-location reviews distinguish these specialized source products from native F12 pools; they do not deny the shared compartments or imply natural splice isoforms. The older same-term action warnings remain documented: the schema consistency_exception field applies to a separate rule table and does not suppress those warnings.
The displayed 615-residue source record and its literal processing features remain unchanged. Signal removal and alpha/beta cleavage chains are processing states, not automatically alternative natural products; 54 Ensembl models likewise do not establish equivalent proteins or functions. No alternative-product IDs are fabricated.
Reference availability flags describe actual complete-original availability, separately from immutable normal-cache metadata and from the narrower text actually read. The two mouse donor papers were acquired by genuine additive normal fetches. The independently fetched human PMID:34967109 cache is reused byte-exact from the direct lane. The provisional union contains 66 normal sources: 38 cited/family dependencies and 28 unused intake sources retained privately, plus the two machine gene files. No normal source was edited or synthesized. Uncited normal records remain private; no source bytes are edited.
F12 encodes a secreted precursor whose activation and further cleavage produce functionally distinct forms. The displayed human UniProt sequence contains 615 residues. Its signal peptide and activated chains are processing features, not splice isoforms. Historical mature-protein residue numbers and experimental fragments are not silently converted to current precursor coordinates. No natural-product equivalence is inferred from a human cell host, a protein-group accession or a yeast fusion construct.
PMID:18725990, Maas et al.: the complete main Results, Methods and Discussion and Figures 1–6 captions were read. Nonfibrillar misfolded aggregates stimulate FXII-dependent prekallikrein activation, whereas the tested amyloid fibrils do not. Chromozym PK measures generated kallikrein, not direct FXII turnover. Surface controls also support FXI activation and coagulation, but the aggregate preparations do not. Plasma activation markers corroborate the contact response. The proposed protective clearance mechanism remains a Discussion model. The functional aggregate assay supports the imported recognition inference without establishing an affinity constant or a structural binding interface. No figure pixels or supplementary figures were inspected.
PMID:34967109, Truong et al.: the accessible complete publisher main was read in Methods 2.1–2.7 and Results 3.1–3.6, including Tables 1–4. Human FXII, α-FXIIa, β-FXIIa, FXI and prekallikrein are explicitly purchased reagents; HRG is isolated from human plasma. The isolated autoactivation assay contains FXII, silica or polyphosphate and zinc, with no added prekallikrein. Generated activity is measured with S-2302 and calibrated against FXIIa. This supports surface-dependent functional autoactivation, separately from reciprocal amplification by kallikrein; it is not a newly inspected cleavage-site sequencing experiment or a measurement of an unstimulated spontaneous rate. Separate FXI and prekallikrein activation assays inhibit FXIIa before reading the respective downstream proteases. SPR and reciprocal immobilization establish HRG binding to α-FXIIa under the tested conditions, with distinct negative results for zymogen and β-FXIIa. Arg353–Val354 is the paper’s numbering. No figure pixels or supplementary material were read. The genuine normal publication record remains abstract-only; this external original supplies the operative evidence.
PMID:6793628, Saito et al.: the complete eight-page original was recovered despite an abstract-only normal record. Methods on pages 1029–1030, Results/Tables II–III, Figures 4–6 captions and Discussion compare normal human plasma Hageman factor with the Washington D.C. patient protein. Controlled trypsin activation, subsequent trypsin blockade, prekallikrein-activator assays, clotting correction and active-site labeling distinguish cleavage from acquisition of enzymatic activity. Added trypsin performs the initial activation in those experiments. No figure pixels were inspected.
PMID:11204562(https://doi.org/10.1055/s-0037-1612914): the identifier and positive endothelial-contact-activation abstract were verified. Complete original Methods and target-reagent controls remain unavailable. Retention of the broad serine-protease function uses independent human enzyme experiments; it does not assert a fresh inspection of the endothelial assay.
PMID:89876, Kaplan et al.: selected actual publisher-original pages 851, 853, 855, 857 and 859 supply Methods, operative Results and captions. Activated Hageman factor fragments activate prekallikrein and FXI upstream of fibrinolysis. Their direct plasminogen-activator effect is negligible at the tested concentrations relative to downstream enzymes. These experiments support zymogen activation and contextual positive regulation of plasminogen activation, without making FXII the principal fibrin-cleaving or plasminogen-activating enzyme. The current GO:0010756 definition permits an increase in activation through this performed upstream step; negligible direct cleavage is not evidence against that regulatory relationship. Complete main access and figure pixels remain unavailable.
PMID:12011093, Isawa et al.: the author-uploaded May 13, 2002 JBC Papers-in-Press version supplies Materials, contact-system assays, SPR Methods and operative Results/Figures 6–9 discussion. Recombinant mosquito hamadarin binds FXII/FXIIa with zinc dependence and inhibits reciprocal contact activation. The human-plasma context is explicit; purchased protein sequence identity is not independently established. The final journal body was not recovered, so manuscript/final equivalence is not assumed. Hamadarin supplies inhibitor activity; that role is not assigned to F12. No pixels were viewed.
PMID:17645545, Isawa et al.: actual publisher-rendered Materials, SPR Methods, operative Results, Figures 4–9 captions and Tables 1–2 were inspected. Separate recombinant triafestin-1/Ti369, triafestin-2/Ti263 and hamadarin experiments support target binding. The paper-numbered FXII 1–77 fragment is not aligned here to the current precursor. The two-state fit is not simplified into an invented one-state dissociation constant. Complete local main access and figure pixels remain unavailable.
PMID:21304106(https://doi.org/10.1182/blood-2010-07-290551): the positive HRG–FXIIa abstract is verified, but the original full Methods and Results remain unavailable. Independent human experiments in PMID:34967109 establish the pair and its α-FXIIa specificity without erasing this earlier source’s access limitation.
PMID:35817195(https://doi.org/10.1016/j.actatropica.2022.106602): the verified abstract reports ayaconin binding to FXII and inhibition when added before activation, rather than inhibition of existing FXIIa. The full assay controls and target reagent provenance remain unread. The interaction remains unresolved rather than being rejected from abstract omissions.
PMID:8710908(https://doi.org/10.1073/pnas.93.16.8552): author-uploaded original Methods and operative Results/Figures 4–5 text show zinc-dependent binding of biotinylated FXII to HUVEC-purified or recombinant gC1qR/C1QBP. The separate cell-antibody experiment measures kininogen binding. C1QBP is not the C1q complement complex, so these assays do not establish complement binding by F12. Complete local main access and figure pixels remain unavailable.
PMID:32814053, Haenig et al.: the complete author-repository paper and Table S2 were recovered. The original Y2H Methods and exact APP–F12 and F12–LAMP2 records were inspected. Table S2 NN1.0 rows 3172 and 5616 contain these pairs; the extended table also retains them. Current IntAct records identify P00748/P05067 and P00748/P13473-2 in array, prey-pooling and validated two-hybrid contexts. The isoform suffix belongs to LAMP2. Human-protein fusion assays in yeast do not establish F12 secretion, activation or exact current insert identity. Study-wide orthogonal assays are not attributed to these specific pairs. APP association is not relabeled as the misfolded-Aβ assay, and LAMP2 association does not establish autophagy. No pixels were viewed.
PMID:25037231(https://pmc.ncbi.nlm.nih.gov/articles/PMC4223627/): complete main Methods/Results and Additional file 2 were read. Sequential extraction produces insoluble ECM-enriched human tissue fractions. Sheets A/B, row 240, contain F12/P00748, including liver detections and the historical group P00748|Q8IZZ5|Q96EF3. Eight total unique peptides and positive normal-liver and liver-metastasis detections support the raw source’s liver context. Compartment-marker controls accompany the sequential extraction; fraction overlap is acknowledged, but target-specific F12 contamination is not demonstrated. The observation does not show that F12 supplies matrix structure or resolve a processed chain. No pixels were viewed.
PMID:23533145(https://pmc.ncbi.nlm.nih.gov/articles/PMC3773505/): complete main Methods/Results, original supplementary protocols and Supplemental Table 2 were read. The P00748/F12 row reports 8/9 unique peptides in cancer-pool analyses and 6/5 in non-cancer-pool analyses. Pooled human EPS-urine undergoes differential ultracentrifugation and two PBS washes of the final pellet; exosome-associated markers are detected. The reported peptide counts are technical analyses of cancer and non-cancer pools, not independent individuals. Soluble/blood-protein carryover is a limitation, not an established explanation for the F12 row. Positive fraction-associated localization is retained without inferring intravesicular topology or an exosome-related molecular function. Processed-chain names in the protein record are not separate natural products. No pixels were viewed.
PMID:2510163(https://pmc.ncbi.nlm.nih.gov/articles/PMC298272/): the identifier and human Washington D.C. variant abstract are verified. The complete original remains inaccessible. The abstract concerns sequence and a proposed catalytic defect, but its lack of an innate immune assay does not refute the imported immune assertion. That context remains unresolved pending an accessible relevant experiment and a defined immune step performed by F12.
The standing campaign instruction explicitly retains supported generic interactions as non-core where no evidence-backed specific molecular function was established, overriding the generic-binding REMOVE default in these eight contexts. A protease-binding term would reverse the role for nonprotease inhibitor partners. These source-specific limits do not negate FXIIa’s established catalytic contact-system role.
The core combines serine-type endopeptidase activity, zymogen activation, intrinsic coagulation, protein autoprocessing and the plasma kallikrein–kinin cascade. The four process terms are nonredundant under current is_a/part_of ancestry. GO:0002353 explicitly includes FXIIa acting on prekallikrein and the downstream kallikrein reaction; its existing accepted coverage is now represented in the same core. Factor XII activation (GO:0002542) is part_of that cascade and is not added as a redundant additional core process. The performed downstream steps are FXI and prekallikrein activation; kallikrein performs kininogen cleavage. The same protease system supports FXII autoactivation, while its initiating molecular species remains a separate question.
The independently read human assays and their exact limits are described above. The core does not infer that a downstream chromogenic readout directly measures FXII turnover, that every aggregate triggers coagulation, or that an FXII-only functional endpoint establishes activity in every initially single-chain molecule. No new figure pixels or supplement reading was undertaken for this synthesis.
Reactome R-HSA-9909048 assigns an F12 active unit in a native receptor-associated contact complex, including the model's explicit single-chain catalytic capacity. R-HSA-9936430 uses extracellular alpha-FXIIa to activate membrane-associated FXI. These routes justify an extracellular and peripheral external-face setting without an integral-membrane assertion, universal bound-state activation or a claim that F12 is the only endothelial prekallikrein activator. The listed complexes organize the catalytic step; no new complex ontology identifier is invented.
The proposed comparisons keep alpha/beta chain processing separate from splice products. They ask whether surface and cofactor composition selects the FXI versus prekallikrein route, and distinguish initial activation from autocatalytic amplification by tracking both cleavage state and active sites.