LPA encodes apolipoprotein(a), a secreted, plasminogen-related glycoprotein whose repeated kringle IV scaffold is covalently disulfide-linked to apolipoprotein B-100 to form the extracellular plasma lipoprotein(a) particle. Kringle IV type 2 is extensively copy-number polymorphic between alleles (2-43 copies); the UniProtKB P08519 represents a 15-copy reference allele, not an alternatively spliced isoform or a universal apo(a) length. Apo(a) kringles mediate binding to lysine-like sites and extracellular partners including fibronectin and apolipoproteins. On fibrin or other surfaces, these interactions can noncatalytically reduce tissue-type plasminogen activator-dependent plasminogen activation. Although apo(a) retains a trypsin-like domain and catalytic triad, its plasminogen-homologous activation junction is defective and recombinant studies found it proteolytically inactive; an older biochemical report of serine-proteinase activity remains disputed in light of the activation-site and recombinant evidence.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0004252 serine-type endopeptidase activity | IEA GO_REF:0000002 | REMOVE | Summary: Removed as a domain-to-GO over-annotation. Apo(a) contains a highly conserved trypsin-like domain and retains the catalytic His-Asp-Ser triad, but its PLG-homologous activation junction is S1819|I1820 rather than the cleavable R580|V581 junction of plasminogen. Reason: The two InterPro signatures establish a trypsin-like fold and an active-site histidine motif, not an activated enzyme. The reproducible sequence analysis found 88.58% identity to the plasminogen domain and retention of H1861, D1904, and S1990, but also replacement of the canonical P1 basic residue at the zymogen activation junction by serine and no UniProt cleavage annotation. This makes activity propagation from domain membership unsafe. It does not resolve the old IDA row, which is reviewed separately with curator deference. Propagation Review Root cause: PROPAGATION BAD Failure modes: PSEUDO OR SUBACTIVITY LOSS Sources checked: InterPro:IPR001254 · Serine proteases, trypsin domain SUPPORTS SOURCE BUT NOT TARGET The domain match establishes fold homology but cannot establish activation or catalysis. InterPro:IPR018114 · Serine protease-like, trypsin family, histidine active site SUPPORTS SOURCE BUT NOT TARGET The active-site signature is retained, but the PLG-like activation junction is not. Supporting Evidence: file:human/LPA/LPA-bioinformatics/RESULTS.md At the homologous apo(a) protease-domain boundary, the sequence is S1819|I1820 and UniProt provides no cleavage annotation. |
| GO:0006508 proteolysis | IEA GO_REF:0000002 | REMOVE | Summary: Removed because this process annotation is propagated solely from protease signatures and inherits the same unresolved activation problem as the inferred serine-type endopeptidase activity. Reason: Participation in proteolysis requires demonstrated peptide-bond hydrolysis in the biological system; a protease-like fold alone is insufficient. Apo(a) retains the catalytic triad but lacks the PLG-homologous basic activation junction, providing a concrete LPA-specific reason not to transfer a generic proteolysis term from these signatures. Propagation Review Root cause: PROPAGATION BAD Failure modes: PSEUDO OR SUBACTIVITY LOSS Sources checked: InterPro:IPR001254 · Serine proteases, trypsin domain SUPPORTS SOURCE BUT NOT TARGET A fold signature does not show that LPA participates in proteolysis. InterPro:IPR018114 · Serine protease-like, trypsin family, histidine active site SUPPORTS SOURCE BUT NOT TARGET Retention of one catalytic motif does not establish an activatable enzyme. Supporting Evidence: file:human/LPA/LPA-bioinformatics/RESULTS.md The S1819 substitution at the homologous PLG activation junction is a concrete reason to treat computationally propagated protease activity as suspect and pseudoenzyme over-annotation as biologically plausible. |
| GO:0034185 apolipoprotein binding | IEA GO_REF:0000117 | ACCEPT | Summary: Accepted as a core binding activity supported by direct interactions with apolipoprotein H and the defining apoB-100 partnership in Lp(a). Reason: Direct interactions with APOH and the covalent apoB-100 partnership that forms Lp(a) establish apolipoprotein binding as a core apo(a) activity. The broad term covers these distinct apolipoprotein partners without implying that every binding event has the same physiological consequence. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00090479 · ARBA rule for apolipoprotein binding SUPPORTS TRANSFER The inferred core binding activity is corroborated by the direct APOH interaction. Supporting Evidence: PMID:9269765 apo(a)/Lp(a) interact in human plasma and in cell culture |
| GO:0005515 protein binding | IPI PMID:2531657 Lipoprotein(a) binds to fibronectin and has serine proteinas... | MODIFY | Summary: Modified from uninformative protein binding to the experimentally supported fibronectin binding term GO:0001968. Reason: The interacting partner P02751 is fibronectin, and the abstract explicitly reports binding of apo(a) to immobilized fibronectin. GO:0005515 discards the biologically useful partner identity, whereas GO:0001968 captures it and is already represented by the source-specific annotation from this study. Proposed replacements: fibronectin binding Supporting Evidence: PMID:2531657 binds to immobilized fibronectin. |
| GO:0005515 protein binding | IPI PMID:9269765 Novel interaction of apolipoprotein(a) with beta-2 glycoprot... | MODIFY | Summary: Modified from generic protein binding to the two informative interactions represented in the normalized source row: apolipoprotein binding and fibronectin binding. Reason: Live QuickGO contains two PMID:9269765 GO:0005515 records, one with APOH (P02749) and one with fibronectin (P02751); normalization correctly combines their WITH/FROM values. The paper reports both an apo(a)-APOH interaction and a fibronectin clone. GO:0034185 and GO:0001968 preserve those partner classes, unlike GO:0005515. Proposed replacements: apolipoprotein binding fibronectin binding Supporting Evidence: PMID:9269765 apo(a)/Lp(a) interact in human plasma and in cell culture PMID:9269765 one as fibronectin. |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-176879 | ACCEPT | Summary: Accepted as the core compartment of secreted apo(a) and extracellular Lp(a) particle assembly. Reason: Reactome places binding of apo(a) to LDL primarily outside cells, on the hepatocyte surface or in blood. This agrees with apo(a)'s signal peptide and well-established plasma residence. Supporting Evidence: Reactome:R-HSA-176879 LPA - LDL binding appears to occur primarily extracellularly in vivo |
| GO:0001968 fibronectin binding | IPI PMID:2531657 Lipoprotein(a) binds to fibronectin and has serine proteinas... | ACCEPT | Summary: Accepted as a directly demonstrated core extracellular binding activity. Reason: The abstract directly reports apo(a) binding to immobilized fibronectin and identifies a fibronectin region contacted by Lp(a). Together with the broader lysine-site-binding properties of apo(a) kringles, this is a directly demonstrated core extracellular binding activity. Supporting Evidence: PMID:2531657 binds to immobilized fibronectin. |
| GO:0004252 serine-type endopeptidase activity | IDA PMID:2531657 Lipoprotein(a) binds to fibronectin and has serine proteinas... | UNDECIDED | Summary: Undecided. The curator-read 1989 study reports apo(a)-localized serine proteinase-type activity, but only its abstract is cached and sequence analysis identifies a noncanonical, plausibly activation-defective junction. Reason: Curator deference precludes removal of this IDA annotation without assessing the full purification, activation, inhibitor, and contaminant-control details. The abstract reports reproducible fibronectin cleavage, peptide-substrate specificity, localization of activity to apo(a), and serine-proteinase-type inhibitor behavior. In tension with that report, P08519 retains the catalytic triad but replaces the PLG R|V activation junction with S|I. The sequence result supports neither acceptance nor confident rejection of the old assay. Supporting Evidence: PMID:2531657 The proteolytic activity of Lp(a) was localized to apo(a) and experiments with inhibitors indicated that the proteolytic activity was of serine proteinase-type. file:human/LPA/LPA-bioinformatics/RESULTS.md This sequence result cannot establish absolute inactivity. |
| GO:0008201 heparin binding | NAS PMID:2531657 Lipoprotein(a) binds to fibronectin and has serine proteinas... | REMOVE | Summary: Removed because the cited abstract shows binding to a heparin-binding domain of fibronectin, not binding of apo(a) to heparin itself. Reason: "Heparin-binding" modifies the C-terminal fibronectin domain used to localize the Lp(a)-fibronectin interaction. No heparin-binding assay or direct apo(a)- heparin interaction is described in the accessible evidence, so this NAS term converts a partner-domain name into an unsupported molecular function. Supporting Evidence: PMID:2531657 The binding of Lp(a) was localized to the C-terminal heparin-binding domain of fibronectin. |
| GO:0034358 plasma lipoprotein particle | IDA PMID:2531657 Lipoprotein(a) binds to fibronectin and has serine proteinas... | ACCEPT | Summary: Accepted as a defining cellular-component annotation: apo(a) is the distinctive protein component of the plasma lipoprotein(a) particle. Reason: The source abstract explicitly describes apo(a) as associated with apoB-100, the main protein component of LDL. Reactome independently models one apo(a) molecule binding an LDL particle to form Lp(a), supporting part_of rather than a transient generic localization. Supporting Evidence: PMID:2531657 the main protein component of low-density lipoprotein. Reactome:R-HSA-176879 forming a Lp(a) lipoprotein particle. |
| GO:0034185 apolipoprotein binding | IPI PMID:9269765 Novel interaction of apolipoprotein(a) with beta-2 glycoprot... | ACCEPT | Summary: Accepted as a directly demonstrated core apolipoprotein-binding activity. Reason: The paper identifies beta-2 glycoprotein I (APOH) in a kringle IV type-2 interaction screen and confirms interaction with apo(a)/Lp(a) by coimmunoprecipitation in human plasma and cell-culture supernatants. This directly supports the core apolipoprotein-binding activity, while the physiological consequence of the APOH interaction remains unresolved. Supporting Evidence: PMID:9269765 apo(a)/Lp(a) interact in human plasma and in cell culture |
| GO:0006629 lipid metabolic process | NAS PMID:8047165 Activation of transforming growth factor-beta is inhibited i... | MARK AS OVER ANNOTATED | Summary: Marked as over-annotated. Apo(a) is a lipoprotein component, but the cited experiment concerns inhibition of plasminogen and TGF-beta activation rather than lipid synthesis, transport, or catabolism. Reason: Membership in a lipid-containing particle does not by itself establish active participation in the broad lipid metabolic process. PMID:8047165 reports vascular effects downstream of inhibited plasminogen activation, so the NAS assertion is biologically adjacent but too broad for the demonstrated mechanism. Supporting Evidence: PMID:8047165 as a consequence of apolipoprotein(a) inhibition of plasminogen activation. |
| GO:0005576 extracellular region | NAS PMID:14718574 The human plasma proteome: a nonredundant list developed by ... | ACCEPT | Summary: Accepted as a core localization supported by detection in the human plasma proteome. Reason: Apo(a) is a secreted plasma protein and a constituent of circulating Lp(a). The cited plasma-proteome compilation supports the broad extracellular-region term without asserting a narrower subcompartment. Supporting Evidence: PMID:14718574 We have merged four different views of the human plasma proteome |
| GO:0004866 endopeptidase inhibitor activity | TAS PMID:8047165 Activation of transforming growth factor-beta is inhibited i... | MODIFY | Summary: Modified to GO:0010757, negative regulation of plasminogen activation. The cited study demonstrates inhibition of a process, not direct stoichiometric inhibition of an endopeptidase by apo(a). Reason: PMID:8047165 reports that apo(a)/Lp(a) inhibits conversion of plasminogen to plasmin and thereby blocks proteolytic TGF-beta activation. That supports negative regulation of plasminogen activation, while GO:0004866 makes the stronger molecular-function claim that apo(a) directly acts as an endopeptidase inhibitor. The accessible abstract does not establish that mode. Because the replacement is a biological-process term, its qualifier must change from `enables` to `involved_in`. Proposed replacements: negative regulation of plasminogen activation Supporting Evidence: PMID:8047165 as a consequence of apolipoprotein(a) inhibition of plasminogen activation. |
| GO:0008015 blood circulation | TAS PMID:8047165 Activation of transforming growth factor-beta is inhibited i... | MARK AS OVER ANNOTATED | Summary: Marked as over-annotated because presence in blood and effects on vascular smooth-muscle signaling do not establish involvement in the process of blood circulation. Reason: The cited transgenic-mouse study concerns inhibition of plasminogen activation, TGF-beta activation, and vascular smooth-muscle-cell proliferation. It does not show that apo(a) carries out or regulates movement of blood through the cardiovascular system. The term therefore conflates a circulating particle and cardiovascular disease relevance with the GO biological process. Supporting Evidence: PMID:8047165 These effects are closely correlated with VSMC activation. |
| GO:0071827 plasma lipoprotein particle organization | IMP PMID:7505444 Cys4057 of apolipoprotein(a) is essential for lipoprotein(a)... | NEW | Summary: Apo(a) is directly required for formation of the Lp(a) particle through its cysteine-dependent association with apoB-100. Reason: Mutation of apo(a) Cys4057 yields secreted free apo(a) without detectable lipoprotein-associated apo(a), while independent site-directed mutagenesis identifies this cysteine in the defining disulfide linkage to apoB-100. These perturbation data support a direct role for LPA in plasma lipoprotein particle organization through formation of the Lp(a) particle. Unlike the narrower assembly term, GO:0071827 does not restrict the organizing interaction to a non-covalent mechanism. Supporting Evidence: PMID:7505444 The same analysis performed with supernatants of cells transfected with plasmids mutated in codon 4057 revealed free apolipoprotein(a) glycoprotein without detectable amounts of lipoprotein-associated apolipoprotein(a). PMID:8366120 Using site-directed mutagenesis, we demonstrated that Cys4057 in apo(a) is involved in disulfide linkage with apoB-100 in Lp(a) particles. |
Loading supporting content…
Download this section (compressed HTML)Q: Under what physiological activation state, if any, can apo(a)'s protease-like domain cleave a substrate despite its noncanonical activation junction and the negative recombinant-protein studies?
Q: How does natural KIV-2 copy number alter the avidity and geometry of apo(a) binding to fibrin, fibronectin, lysine-like sites, and apolipoprotein partners?
Q: Does APOH binding measurably change Lp(a) assembly, extracellular retention, uptake, or antifibrinolytic activity in human plasma?
Experiment: Purify sequence-defined apo(a) and covalent Lp(a) particles under protease-free conditions, compare wild type with activation-junction and catalytic-serine mutants, verify processing by intact-mass spectrometry and N-terminal sequencing, and test fibronectin plus orthogonal peptide substrates with active-site probes and contaminating-protease spike-in controls.
Hypothesis: Physiological apo(a) is activation-defective and lacks intrinsic serine-endopeptidase activity.
Type: biochemical protease reconstitution and active-site profiling
Experiment: Produce an allelic series of otherwise matched recombinant apo(a)/Lp(a) particles with defined KIV-2 copy counts, then measure fibrin, degraded-fibrin, fibronectin, and lysine-analogue binding together with tPA-dependent plasmin generation on each surface.
Hypothesis: Apo(a) KIV-2 copy number quantitatively changes multivalent extracellular binding and surface-dependent inhibition of plasminogen activation.
Type: allele-resolved binding and fibrinolysis assay
Experiment: Compare full-length apo(a), a protease-domain deletion, and precise KIV-10 lysine-site mutants in matched Lp(a) particles using fibrin-clot plasminogen activation and lysis assays; confirm equivalent APOB coupling and particle abundance before functional comparison.
Hypothesis: The KIV-10 strong lysine-binding site is required for apo(a)'s surface-dependent antifibrinolytic effect, whereas the protease-like domain is dispensable.
Type: domain-resolved structure-function analysis
Experiment: Immunodeplete APOH from human plasma and restore it at physiological concentration, then quantify Lp(a) assembly state, fibronectin/fibrin binding, plasminogen activation, and receptor-mediated uptake with an apo(a) KIV-2 interaction-site mutant as a specificity control.
Hypothesis: APOH binding modulates a specific extracellular property of Lp(a) rather than merely reporting a detectable physical interaction.
Type: plasma depletion and add-back functional assay
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: Whether apo(a) ever acquires intrinsic serine-endopeptidase activity in a physiological context remains unresolved.
OPEN BIOLOGY MF_DARK
What is known: The protease-like domain retains the His-Asp-Ser triad, but the canonical plasminogen activation-site arginine is replaced by serine, recombinant apo(a) remained inactive even after an engineered activation-site restoration, and only one older curator-read study reports apo(a)-localized proteolysis.
Significance: Resolving the conflict determines whether the experimental protease annotation represents a rare context-dependent activity or an assay-specific attribution; current evidence does not justify making catalysis a core function.
What would resolve it: Reconstitute purified, sequence-defined apo(a) and Lp(a) preparations with activation-state controls, orthogonal substrates, active-site labeling, and protease-contamination controls.
Provenance (the field's own admissions):
Gap: The relative physiological contributions of individual apo(a) kringle interactions to Lp(a) deposition and antifibrinolytic activity are not defined across the natural range of KIV-2 copy-number alleles.
OPEN BIOLOGY
What is known: Strong and weak lysine-binding sites have been measured in isolated kringles, and a lysine-binding-site variant abolishes lysine-Sepharose binding, while natural alleles vary specifically in KIV-2 repeat count.
Significance: Allele-specific avidity could connect apo(a) size polymorphism to particle localization and inhibition of plasminogen activation without invoking different splice isoforms.
What would resolve it: Compare matched recombinant apo(a)/Lp(a) particles spanning KIV-2 copy number and targeted kringle lysine-site mutations in quantitative fibrin, fibronectin, and plasminogen-activation assays.
Gap: The biological consequence of the directly observed apo(a)-APOH interaction is unknown.
OPEN BIOLOGY
What is known: APOH was identified with KIV-2 and coimmunoprecipitated with apo(a)/Lp(a) in human plasma and cell-culture supernatants, but the accessible study does not establish a downstream molecular or physiological effect.
Significance: Establishing an effect would clarify how this core apolipoprotein-binding activity influences Lp(a) biology beyond the demonstrated physical contact.
What would resolve it: Test whether APOH changes Lp(a) assembly, fibrin/fibronectin binding, plasminogen activation, or cellular uptake using APOH depletion and add-back.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)