CTSD

Summary: The established aspartic endopeptidase mechanism is more informative than this broad activity.

Reason: The mouse Ctsd/Ensembl source is a genuine orthology-based assertion, and the human enzyme retains the catalytic Asp97/Asp295 pair. Human sequence and native/pepstatin structural evidence support the specific aspartic endopeptidase mechanism. Refine the broad positive activity without altering its original evidence or donor tuple. The complete structural abstract was read; full original structural Methods were not recovered.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The aspartic endopeptidase IBD lies on the verified human CTSD path.

Reason: The actual PTHR47966 tree contains human P07339 at PTN002492008 with an exact 412-residue MSA match. The GO:0004190 IBD is at ancestral PTN002492107 on that path, with the literal descendant sources preserved. Conserved human catalytic aspartates and positive biochemical evidence support retention; no target-specific loss or divergence is indicated. The deeper enzyme-family placement does not transfer the specialized physiological functions of renin or other descendants.

Summary: Independent sequence, family, EC and mouse evidence agree on aspartic endopeptidase activity.

Reason: The literal IPR001461/IPR001969/IPR012848/IPR033144 and EC:3.4.23.5 mappings agree with the human active-site residues and structural classification. The conserved mouse source supplies corroboration. Preserve the combined-source assertion and its complete source list.

Summary: Human cathepsin D is an established aspartic endopeptidase.

Reason: The complete 3927292 abstract identifies the human precursor and its mature-region relationship to aspartyl proteases. Independent human native/pepstatin structural evidence and biochemical studies corroborate the specific activity. This is a positive TAS assertion; unavailable full Methods are not grounds to reject it.

Summary: A mixed Reactome catalyst set incorrectly assigns a cysteine mechanism to CTSD.

Reason: Both modeled reactions use catalyst 2130638, whose cysteine-specific GO:0004197 activity is attached to mixed cathepsin CandidateSet 2130682. Its actual hasMember 1181223 resolves to human P07339. CTSD instead uses the aspartic mechanism, with catalytic Asp97/Asp295 and independent human structural evidence. The incompatible nucleophile assignment is a demonstrated mapping error; this does not deny general antigen proteolysis, remove the positive location assertion, or rewrite the original TAS source.

Summary: A mixed Reactome catalyst set incorrectly assigns a cysteine mechanism to CTSD.

Reason: Both modeled reactions use catalyst 2130638, whose cysteine-specific GO:0004197 activity is attached to mixed cathepsin CandidateSet 2130682. Its actual hasMember 1181223 resolves to human P07339. CTSD instead uses the aspartic mechanism, with catalytic Asp97/Asp295 and independent human structural evidence. The incompatible nucleophile assignment is a demonstrated mapping error; this does not deny general antigen proteolysis, remove the positive location assertion, or rewrite the original TAS source.

Summary: Retain the positive amyloid-beta enzyme–substrate association.

Reason: The complete abstract explicitly reports cathepsin D cleavage of amyloid-beta peptide and aluminum-dependent inhibition. Two complete positive MINT records join the original PMID to P07339 and P05067, with literal feature ranges 1–40 and 1–42. These represent the tested amyloid-beta peptides, not residues 1–40 or 1–42 of the APP precursor. This supports enzyme–substrate association rather than a separate core binding function or an interaction with full-length APP. The abstract does not identify the cathepsin reagent species; human participant assignment is retained from the curated records, not invented from unavailable Methods.

Summary: Retain the measured CTSD–progranulin association.

Reason: The actual original and supplementary Methods describe human CTSD residues 19–end with a secretion signal/tag, human progranulin constructs, and co-immunoprecipitation in HEK293 or mouse N2a contexts. The interaction and effects on CTSD activity are positive; these lysate/media assays are not purified binary binding. CTSD is the regulated protease, not an enzyme activator merely because progranulin increases its activity.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the positive association with the specified processed granulin product.

Reason: The original supplementary experiments test individual GFP-tagged granulins and FLAG-tagged human CTSD, with positive co-immunoprecipitation of all seven granulin constructs at differing strengths. Preserve this row's literal PRO product identifier separately from full-length GRN and the other processed products. These are cell-derived tagged preparations, not proof of a common affinity or a purified direct interface for every granulin.

Summary: Retain the source-specific positive EPN3 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and EPN3. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The unreviewed I6L9I8 clone is a 318-residue epsin-3 product. A more specific functional binding class is not established for this assayed clone. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: The positive WWP2 association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 870-residue WWP2 product retains its HECT E3-ligase region. Ubiquitin protein ligase binding describes the positive partner class; it does not establish ubiquitination of CTSD or ligase activity by CTSD.

Proposed replacements: ubiquitin protein ligase binding

Summary: Retain the source-specific positive NPFF association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and NPFF. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The tested NPFF isoform-2 precursor is not an experimentally identified mature neuropeptide. Preserve the precursor association without substituting neuropeptide-receptor binding. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain binding to the putative P09565 product without assigning an IGF2 function.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The current official P09565 record is a 113-residue putative insulin-like growth factor 2-associated protein with no assigned gene symbol. It is not IGF2 itself, and neither growth-factor binding nor a better established partner class follows from its name. The positive generic interaction remains contextual.

Summary: The positive TEAD1 association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 357-residue TEAD1-2 backbone retains the complete TEA DNA-binding region. Its 58-residue deletion affects the broader activation region, so equal transcriptional output across isoforms is not assumed. The official sequence-specific Pol II transcription-factor class and retained defining DNA-binding region support this partner-class refinement; the CTSD interaction does not demonstrate transcriptional regulation.

Proposed replacements: RNA polymerase II-specific DNA-binding transcription factor binding

Summary: The positive VDAC2 association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 294-residue P45880 backbone is the official displayed VDAC2 isoform 3, including its complete channel sequence. Transmembrane transporter binding describes that supported partner class; the isoform number does not make this a fragment. No mitochondrial localization, ion transport or scramblase activity is assigned to CTSD.

Proposed replacements: transmembrane transporter binding

Summary: The positive histone H3.1 association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. P68431 is histone H3.1. Histone binding is a supported class refinement of the positive interaction, without asserting endogenous chromatin residency or histone proteolysis by CTSD.

Proposed replacements: histone binding

Summary: The positive KIF1A association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The Q12756 partner is kinesin KIF1A. The positive interaction supports kinesin binding without assigning motor activity, transport regulation or endogenous cargo attachment to CTSD.

Proposed replacements: kinesin binding

Summary: Retain the source-specific positive WTAP association.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 151-residue WTAP-2 product is a short form of a regulatory component of the methyltransferase complex, not the catalytic methyltransferase. Neither the gene-level role nor this construct establishes a more specific enzyme-binding class; positive generic binding is retained.

Summary: Retain the source-specific positive LCE2A association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and LCE2A. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The exact LCE2A product is a late cornified-envelope protein. No more specific experimentally justified molecular binding class is established here. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: The retained ZDHHC20 catalytic architecture supports an inferred enzyme-binding refinement.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 320-residue ZDHHC20-4 product retains residues 1–314, including the four transmembrane segments, DHHC active site, TTXE motif and PaCCT region. The original human ZDHHC20 structural and activity study maps the resolved catalytic architecture to residues 5–299, upstream of the isoform-specific replacement at 315–365. This supports broad enzyme binding at an explicitly inferred partner-class scope. The study assayed full-length ZDHHC20, not this splice product; equal kinetics and a physiological CTSD acylation mechanism are not claimed.

Proposed replacements: enzyme binding

Summary: Retain the source-specific positive ZNF497 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 4 complete positive records for CTSD and ZNF497. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The Q6ZNH5 record suggests a possible transcriptional role but does not establish a DNA-binding-transcription-factor class. A separate human-cell LUMIER record corroborates the interaction without supplying that missing class evidence. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: The positive GPR141 association supports a more informative partner-class term.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. Q7Z602 is the full-length officially classified orphan receptor GPR141. The actual signaling receptor binding definition does not require occupation of a ligand site or a change in receptor activity; receptor ligand activity is a separate term. This positive partner-class refinement therefore does not imply that CTSD is an agonist, that the interface is extracellular, or that endogenous GPCR signaling was measured.

Proposed replacements: signaling receptor binding

Summary: Retain the source-specific positive MOGAT3 association.

Reason: The three literal positive records describe array, pooled and validated yeast two-hybrid assays for the same CTSD–partner pair, rather than three independent physiological studies. The accession-specific backbone was reconstructed, but tags, native processing, colocalization and protease-dependent function are not established by those sequence identifiers. The 344-residue MOGAT3-3 product retains the transmembrane and Cys265-containing regions but replaces residues 291–341 with a distinct tail. The consulted evidence does not delimit a complete functional catalytic unit that excludes that replaced region, unlike the mapped ZDHHC20 catalytic architecture. Positive generic binding is retained while this product-specific enzyme-class inference remains unresolved; no inactivity or nonbinding is inferred.

Summary: Retain the source-specific positive SUMF2 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and SUMF2. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. SUMF2 is catalytically inactive rather than the sulfatase-modifying enzyme SUMF1. Preserve its positive association without an enzyme-binding assignment based on the family name. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive PGLYRP3 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and PGLYRP3. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. PGLYRP3 has peptidoglycan-recognition and antibacterial functions. The pair does not establish a signaling-receptor role or transfer its antibacterial activity to CTSD. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive SYT3 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and SYT3. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The synaptotagmin-3 association does not by itself establish calcium sensing or membrane-fusion regulation by CTSD. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive SPRY4 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and SPRY4. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. SPRY4 is a signaling regulator, not a protein kinase. The positive pair is retained without kinase-binding or pathway-regulation claims. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive EPS8L2 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and EPS8L2. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The EPS8-like partner is a scaffold/regulatory protein; this assay does not show intrinsic guanine-nucleotide exchange activity or justify an enzyme-binding label. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive DLGAP2 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and DLGAP2. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The Q9P1A6-3 scaffold product is preserved literally. Its name does not establish binding to a purified PDZ domain, and no such domain-specific assay is substituted for this record. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: Retain the source-specific positive IRAG1 association.

Reason: The original Haenig neurodegenerative-disease interactome citation joins 3 complete positive records for CTSD and IRAG1. Human protein constructs were tested predominantly by yeast two-hybrid; backbone sequence hashes and recorded features were reconstructed independently. The Q9Y6F6-3 backbone maps exactly, while the literal tested feature is residues 207–803. This fragment association does not establish CTSD regulation of calcium release or PKG activity. Physiological colocalization and protease-dependent function are not established by this screen.

Summary: The colostrum study is accessible only at abstract scope, without the CTSD-specific target table.

Reason: The abstract describes 151 minor proteins in aqueous human colostrum but does not identify the CTSD observation. Extracellular CTSD is independently well supported; that general fact does not verify this particular proteomics record. Retain the HDA source pending its target row rather than rejecting it because the abstract foregrounds a proteome.

Summary: Human aortic proteomics directly detects extracellular-associated CTSD.

Reason: The actual target table identifies CATD in the 0.5 M NaCl extract, with five unique peptides. The study separates soluble extracellular-associated material from more tightly retained matrix fractions. This supports extracellular occurrence, not CTSD as a structural matrix constituent or proof of secretion by a particular cell type.

Summary: CTSD is detected in the human venous extracellular-matrix preparation.

Reason: The target CATD row occurs in the guanidine fraction of normal saphenous veins in the original normal/varicose comparison. Sequential NaCl/SDS/guanidine extraction and mass spectrometry support the contextual localization. Differential detection is not itself a causal role in venous hypertension.

Summary: The extracellular-region IBD is correctly inherited but describes a context outside the principal lysosomal role.

Reason: The exact GO:0005576 IBD lies at vertebrate PTN008723784 on the P07339 path, with mouse Ctsd experimental grounding and matching literal source IDs. The mouse PMID19710420 abstract concerns hydrolase secretion without naming the target; independent human extracellular proteomics and plaque localization corroborate the biology. No donor-count or self-evidence objection is made.

Summary: Mouse evidence and the UniProt secretion mapping support extracellular occurrence.

Reason: The original mouse/Ensembl and SL-0112 sources are retained. Human vascular proteomics and extracellular plaque-associated CTSD support this nonexclusive location, which does not displace its dominant endolysosomal activity.

Summary: The pig extracellular-matrix source cannot yet be adjudicated at the target-table level.

Reason: The actual donor is porcine P00795 and the cited ischemia/reperfusion study concerns pig cardiac extracellular matrix. Its available abstract does not identify the cathepsin D target row; the present donor snapshot lacks a matching experimental assertion but that absence is not evidence of an erroneous historical ISS. Preserve the source and leave its specific experimental grounding unresolved.

Summary: The CTSD-specific plasma-proteome entry remains unverified.

Reason: The complete abstract describes an integrated list of 1175 plasma proteins, but the CTSD entry in its component datasets was not recovered. Extracellular CTSD is biologically plausible and independently supported; the source-specific NAS assertion remains uncertain rather than being declared incorrect from an abstract-only review.

Summary: The actual endostatin reaction models extracellular processed CTSD.

Reason: The reaction links catalyst set 2471626 to CTSD complex 2471622, composed of light/heavy processed chains derived from P07339 in extracellular space. The source labels this assembly a dimer, but its two components are proteolytic chains, not proof of a homodimer of intact CTSD precursors. This location assertion does not adopt the separate serine-type activity attached to the mixed reaction catalyst.

Summary: The actual tertiary-granule exocytosis model contains extracellular CTSD.

Reason: The full event 6798745 input/output sets resolve human P07339 rather than relying on a pathway title. The extracellular output set 6806432 contains CTSD entity 2990853. This supports a modeled release context without making CTSD an exocytosis motor or transferring other members' catalytic functions.

Summary: The actual specific-granule exocytosis model contains extracellular CTSD.

Reason: The full event 6798749 input/output sets resolve human P07339 rather than relying on a pathway title. The extracellular output set 6806312 contains CTSD entity 2990853. This supports a modeled release context without making CTSD an exocytosis motor or transferring other members' catalytic functions.

Summary: The actual ficolin-rich-granule exocytosis model contains extracellular CTSD.

Reason: The full event 6800434 input/output sets resolve human P07339 rather than relying on a pathway title. The extracellular output set 6806481 contains CTSD entity 2990853. This supports a modeled release context without making CTSD an exocytosis motor or transferring other members' catalytic functions.

Summary: Retain broad cytoplasmic localization as a contextual location.

Reason: ARBA00026971 branch 603 matches the recorded FunFam predicates. Cytoplasm includes cytoplasmic organelles, so the established endolysosomal pools are compatible with this broad location. The more informative lysosomal and endosomal lumen annotations define the principal localization; this broad electronic assertion is retained as non-core and does not imply constitutive soluble cytosolic residence.

Summary: The lysosome IBD and human experiments independently support the principal compartment.

Reason: The GO:0005764 IBD is at PTN008723784 on the exact human tree/MSA path. Mouse, human and other descendant experimental sources are literal and legitimate, including the target's own grounding. Human localization and human-cDNA targeting experiments corroborate the ancestral placement; no loss is evident.

Summary: The original apoptosis experiment confirms endogenous lysosomal CTSD.

Reason: Human AG1518 fibroblast immunocytochemistry shows endogenous CTSD in lysosomes before manipulation. The later microinjection/leakage context is distinct from this baseline compartment. The injected Sigma reagent species is unstated and is not relabeled as purified human CTSD.

Summary: The human brain study explicitly localizes enzymatically active CTSD.

Reason: The complete target-specific abstract reports lysosomal proteases, including cathepsin D, in intracellular lysosomal and extracellular plaque contexts in Alzheimer brain. Retain the positive lysosome annotation without making the pathological distribution its sole normal function; original full Methods were not recovered.

Summary: The positive curator lysosome assertion agrees with independently demonstrated CTSD localization.

Reason: The source is a cystinosin-isoform study, and its accessible abstract does not identify the CTSD control experiment. That title is not evidence of a wrong-gene annotation. Retain curator deference for the well-established CTSD lysosome location, corroborated by the independently inspected CTSD localization controls in PMID19723497, while leaving this original assay detail unverified.

Summary: Retain the lysosomal hydrolase localization at its documented human-brain scope.

Reason: The original abstract concerns abnormal distribution of multiple lysosomal hydrolase classes in Alzheimer brain but does not expose all target-specific images. The positive curated CTSD localization agrees with the separately explicit human CTSD studies; no title-based experimental rejection is warranted. Source-specific image attribution remains unverified.

Summary: The actual NPC2 study uses CTSD as a lysosomal localization and fractionation control.

Reason: Selected original HeLa Methods/Results identify CTSD in lysosomal confocal and biochemical-fractionation controls while examining NgBR/NPC2. This directly supports location; it does not transfer the cholesterol-trafficking activity of the experimental target to CTSD.

Summary: The human Alzheimer-brain abstract explicitly names CTSD lysosomal localization.

Reason: CTSD and other lysosomal-proteinase antigens occur in neuronal lysosomes and plaque-associated degenerating perikarya. Preserve the source-specific location without asserting that plaque association is constitutive in normal brain; complete original figures were unavailable.

Summary: Concordant mouse, family and UniProt mappings support lysosomal localization.

Reason: The full original P18242/Ensembl, IPR033144 and SL-0158 source list is preserved. These mappings agree with human CTSD localization and processing, rather than substituting family membership alone for compartment evidence.

Summary: The lysosomal membrane-specific localization is not resolved by the recovered source.

Reason: The complete abstract directly establishes that human mutant CTSD expressed in rodent cells reaches lysosomes and remains catalytically active as a single chain. It does not distinguish lumenal residency from the membrane-specific association asserted here, and full original Methods/Results were not recovered. A soluble enzyme can associate with a membrane; no categorical exclusion or term replacement is inferred.

Summary: Cytosolic CTSD is supported in a perturbed apoptosis context.

Reason: The original microinjection experiment introduces active cathepsin D into human fibroblast cytosol and later observes endogenous lysosomal leakage. The injected Sigma reagent species is unstated. Retain the experimentally induced/secondary redistribution context without describing cytosol as the principal constitutive compartment.

Summary: The deep proteolysis IBD is supported by the verified family path and CTSD chemistry.

Reason: The exact GO:0006508 IBD at PTN002492107 lies on the human path. Complete descendant queries preserve their multiplicity, and experimentally annotated proteolytic child processes can ground the broader process even where an exact current parent row is absent. Human self-evidence among this row's sources is legitimate. CTSD itself performs peptide-bond cleavage, so its participation is catalytic rather than a substrate-necessity inference.

Summary: The source explicitly measures cathepsin D proteolysis of human apoE substrates.

Reason: The full abstract reports cleavage of both recombinant lipid-free human apoE and lipidated human plasma apoE, with pepstatin sensitivity. The enzyme reagent species is not stated in the abstract, so the positive human annotation is retained with curator deference rather than an invented purified-human-enzyme claim.

Summary: Aspartic-protease family signatures agree with performed CTSD proteolysis.

Reason: The four literal InterPro mappings describe a catalytically competent aspartic protease and agree with the human active-site features and cleavage assays. This broad process is directly performed by CTSD rather than inferred solely from a phenotype.

Summary: The human CTSD single-chain mutant remains an active protease.

Reason: The complete abstract establishes human mutant CTSD expressed in rodent cells, correct lysosomal targeting, pepstatin binding and enzymatic activity despite failure of terminal double-chain maturation. This supports proteolysis and shows that processing into two chains is not required in that assay.

Summary: The positive peptide-cleavage activity can be specified as aspartic endopeptidase activity.

Reason: The source explicitly reports apoE cleavage and pepstatin sensitivity; it does not identify the enzyme reagent species in the accessible abstract. Human active-site and structural evidence establish the specific mechanism. Original full Methods were unavailable, but the refinement does not reject the curator's positive experimental result.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The established aspartic endopeptidase mechanism is more informative than this broad activity.

Reason: The mouse Ctsd/Ensembl source is a genuine orthology-based assertion, and the human enzyme retains the catalytic Asp97/Asp295 pair. Human sequence and native/pepstatin structural evidence support the specific aspartic endopeptidase mechanism. Refine the broad positive activity without altering its original evidence or donor tuple. The complete structural abstract was read; full original structural Methods were not recovered.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The positive peptide-cleavage activity can be specified as aspartic endopeptidase activity.

Reason: The source explicitly reports active human single-chain mutant CTSD in rodent cells, not a splice isoform or a requirement for terminal two-chain processing. Human active-site and structural evidence establish the specific mechanism. Original full Methods were unavailable, but the refinement does not reject the curator's positive experimental result.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The endosomal membrane-specific localization is not resolved by the recovered source.

Reason: The complete abstract directly establishes that human mutant CTSD expressed in rodent cells reaches lysosomes and remains catalytically active as a single chain. It does not distinguish lumenal residency from the membrane-specific association asserted here, and full original Methods/Results were not recovered. A soluble enzyme can associate with a membrane; no categorical exclusion or term replacement is inferred.

Summary: The established aspartic endopeptidase mechanism is more informative than this broad activity.

Reason: The mouse Ctsd/Ensembl source is a genuine orthology-based assertion, and the human enzyme retains the catalytic Asp97/Asp295 pair. Human sequence and native/pepstatin structural evidence support the specific aspartic endopeptidase mechanism. Refine the broad positive activity without altering its original evidence or donor tuple. The complete structural abstract was read; full original structural Methods were not recovered.

Proposed replacements: aspartic-type endopeptidase activity

Summary: Retain contextual antigen processing with CTSD-specific evidence and limits.

Reason: The Reactome pathway places CTSD in a mixed endolysosomal catalyst set; this does not establish CTSD-specific lip10 cleavage, and the separately named initial-aspartic event resolves to LGMN. Independent PMID1328388 explicitly reports that purified cathepsin D generates ovalbumin material recognized by MHC-II-restricted T cells on fixed antigen-presenting cells; its abstract was read, while enzyme provenance and full assay details remain unverified. In the complete relevant original text of PMID9539769, mouse CatD deficiency modestly changes presentation of selected epitopes but leaves overall presentation and invariant-chain degradation intact. Together these findings support a contextual contribution to antigen processing, without an essential role, a human-specific assay claim, or a new antigen-presentation core.

Summary: The human aortic target table supports extracellular-matrix-associated CTSD.

Reason: Actual CATD peptides are present in the aortic extracellular extraction workflow, including the 0.5 M NaCl-associated fraction. This supports the broad annotation at an extracellular-associated proteome scope, not stable structural incorporation or a demonstrated matrix-remodeling reaction by CTSD in this experiment.

Summary: The genuine colon/liver ECM workbook contains the human CTSD target.

Reason: Supplementary Table S1 row 205 identifies P07339/CTSD among matrisome-associated ECM regulators in the human normal-colon, colon-tumor and liver specimens. The exact complete-dataset and matrisome sheets repeat the same underlying target evidence rather than independent experiments. Sequential enrichment and target peptides support ECM association, not proof of an exclusively insoluble matrix protein or causal metastasis activity.

Summary: The actual myeloma bone-marrow workbooks verify CTSD in the ECM preparation.

Reason: Supplementary S2 row 29 and S3 complete-dataset row 93 identify human P07339/CTSD, with group-specific observations retained separately. The two tables use different detection summaries and are not forced into identical group calls. Target-level workbook evidence supports the curated localization; full original Methods were not recovered and no myeloma-specific CTSD mechanism is asserted.

Summary: The human breast/omentum study has explicit CTSD peptide records in its ECM workflow.

Reason: The genuine supplementary workbook contains 11 target measurement rows across breast and omentum fractionation sheets, plus two derivative hydroxylation summaries. Crude ECM, soluble material and prefractionation approaches are distinguished. These positive target peptides support contextual ECM association, not a universal insoluble-matrix structural role or a count of 13 independent experiments.

Summary: Endosomal cathepsin D activity supports retention.

Reason: The mouse donor study measures cathepsin D activity in isolated liver endosomes. Prior endosomal fractionation and soluble-extract work in PMID11779865 supports luminal enzyme activity. Preserve the exact mouse-derived transfer and the nonexclusive localization.

Summary: The specific-granule input set contains human CTSD.

Reason: Reaction 6798749 has input set 6799354 containing P07339-derived entity 6801272 in specific-granule lumen and an extracellular output containing entity 2990853. Preserve this modeled secretory-granule context without making it the dominant location or assigning secretion machinery activity to CTSD.

Summary: Return of the receptor to a signal-ready plasma-membrane state remains unresolved.

Reason: PMID27703010 measures enzyme activity and receptor distribution/phosphorylation. PMID11779865 tests endosomal receptor abundance and phosphorylation after pepstatin, rather than resolving return to the signal-ready plasma membrane. These results leave CTSD involvement in the specific recycling process uncertain. UNDECIDED also covers unresolved interpretation of accessible experiments; it is not restricted to inaccessible papers. This judgment neither requires CTSD to be a transport motor nor asserts that receptor recycling is absent.

Summary: ApoE cleavage supports a substrate-specific lipoprotein-catabolic role.

Reason: The complete original abstract explicitly reports cathepsin D cleavage of lipidated human plasma apoE as well as lipid-free apoE, with pepstatin inhibition. CTSD performs the cleavage rather than merely being consumed by this process. Reagent species and full Methods remain unresolved, and the disease-pathogenesis proposal is not treated as a demonstrated causal outcome.

Summary: Retain the UniProt melanosome mapping as a contextual proteomic location.

Reason: The actual SL-0161 source records melanosomal occurrence, consistent with lysosome-related organelle biology. The underlying melanosome proteomics abstracts do not expose the CTSD target table, so no new direct imaging claim or melanin-synthesis activity is inferred. The supported principal endolysosomal location is compatible with this curated electronic context. The current UniProt subcellular-location statement explicitly records mass-spectrometric identification in melanosome fractions from stages I–IV; it is the checkable source for this electronic mapping, not a newly verified imaging experiment.

Summary: Active cytosolic cathepsin D can promote caspase-dependent apoptosis.

Reason: The original human-fibroblast experiment uses active-enzyme microinjection, inactive/pepstatin controls and time-resolved cytochrome-c redistribution preceding caspase activation. It supports an upstream pro-apoptotic contribution in that experimental context, not a normal universal death program or a known direct BID-cleavage mechanism. Sigma reagent species is unstated.

Summary: The actual CTSD entity in the reaction is located in lysosomal lumen.

Reason: Reaction 2130349 resolves through catalyst 2130638 and mixed set 2130682 to human P07339 entity 1181223, with lysosomal-lumen location. The incorrect cysteine-specific activity on the mixed set is a separate assertion; it does not invalidate CTSD's experimentally established lumenal location.

Summary: The actual CTSD entity in the reaction is located in lysosomal lumen.

Reason: Reaction 2130706 resolves through catalyst 2130638 and mixed set 2130682 to human P07339 entity 1181223, with lysosomal-lumen location. The incorrect cysteine-specific activity on the mixed set is a separate assertion; it does not invalidate CTSD's experimentally established lumenal location.

Summary: The estrogen-responsive expression model places the CTSD product in lysosomal lumen.

Reason: The model output 9023844 contains the processed heavy/light chains 9023737 and 9023739 derived from P07339. Its dimer name denotes two chains in this representation, not an experimentally established homodimer of full precursors. CTSD is the expressed enzyme product, not the transcriptional regulator of this reaction.

Summary: CTSD is detected in the tested detergent-resistant membrane fraction.

Reason: The actual HUVEC SILAC table identifies CATD/P07339 after detergent-resistant-membrane sucrose fractionation in the flotillin-1 perturbation study. This supports the source-specific raft-associated biochemical fraction, with fractionation limits; it does not establish an integral-membrane topology or assign TLR3 signaling to CTSD.

Summary: The established aspartic endopeptidase mechanism is more informative than this broad activity.

Reason: The mouse Ctsd/Ensembl source is a genuine orthology-based assertion, and the human enzyme retains the catalytic Asp97/Asp295 pair. Human sequence and native/pepstatin structural evidence support the specific aspartic endopeptidase mechanism. Refine the broad positive activity without altering its original evidence or donor tuple. The complete structural abstract was read; full original structural Methods were not recovered.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The established aspartic endopeptidase mechanism is more informative than this broad activity.

Reason: The explicit mouse Ctsd P18242 source is a genuine curator-transferred orthology assertion, and the human enzyme retains the catalytic Asp97/Asp295 pair. Human sequence and native/pepstatin structural evidence support the specific aspartic endopeptidase mechanism. Refine the broad positive activity without altering its original evidence or donor tuple. The complete structural abstract was read; full original structural Methods were not recovered.

Proposed replacements: aspartic-type endopeptidase activity

Summary: The author urinary-exosome database explicitly identifies CTSD.

Reason: The genuine NHLBI target table lists cathepsin D preproprotein, CTSD, NP_001900 and one peptide, linked by reference 2 to PMID19056867. Retain the positive source-specific HDA occurrence with its single-peptide and preparation limits; it does not imply that vesicle formation is CTSD's function.

Summary: The separate expressed-prostatic-secretion exosome target record remains unverified.

Reason: The scientific main text and isolation/proteomics workflow were read, but its CTSD-specific supplementary target row was not recovered. Prior files named as spreadsheets contain HTML and are not treated as target data. The independently verified PMID19056867 urinary-exosome row does not verify this different source assertion. Preserve the HDA record pending the actual table.

Summary: The original experiment supports broad regulation of cytochrome-c redistribution.

Reason: Microinjected active cathepsin D induces mitochondrial-to-cytosolic cytochrome-c redistribution before effector-caspase activity in human fibroblasts. The broad regulatory annotation is retained without identifying a direct substrate or claiming direct mitochondrial-membrane action. The tempting GO:0090200 refinement explicitly requires DIRECT positive regulation and is not justified by these controls; BID cleavage remains a hypothesis.

Summary: The paper demonstrates upstream induction rather than execution-phase work by CTSD.

Reason: The measured order is cytochrome-c redistribution followed by caspase activation and later apoptotic morphology. These results support positive regulation of apoptosis, but do not demonstrate CTSD performing the execution-phase dismantling asserted here. This is a source-specific stage distinction, not a categorical claim that cathepsins cannot contribute to execution in another system.

Summary: Retain insulin catabolism as a demonstrated substrate-specific role.

Reason: The 2016 activity-kit experiment is indirect for insulin cleavage. Its cited PMID11779865 directly tests rat endosomal immunodepletion and purified bovine enzyme; matching human cathepsin D digestion is reported without displayed data. This supports the biological process while preserving the original mouse-transfer provenance and the separate uncertainty about receptor recycling.

Summary: The actual model includes CTSD in tertiary-granule lumen.

Reason: The input set 6800171 of event 6798745 contains CTSD entity 6801276, resolved to human P07339 and the specified granule compartment. This is a positive model-specific secretory context, kept separate from constitutive lysosomal proteolysis and from the proteins that execute exocytosis.

Summary: The actual model includes CTSD in ficolin-1-rich-granule lumen.

Reason: The input set 6800431 of event 6800434 contains CTSD entity 6801321, resolved to human P07339 and the specified granule compartment. This is a positive model-specific secretory context, kept separate from constitutive lysosomal proteolysis and from the proteins that execute exocytosis.