Dermatan-sulfate epimerase (DSE, also called SART2) is a membrane-associated glycoprotein of the secretory pathway that converts glucuronic acid residues in chondroitin chains into iduronic acid. This modification generates the iduronic-acid-containing regions of dermatan sulfate proteoglycans. Its manganese-dependent catalytic domain is N-glycosylated, and the protein is associated with endoplasmic-reticulum and Golgi compartments. The resulting dermatan-sulfate structures contribute to extracellular-matrix organization.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000139 Golgi membrane | EXP PMID:10679095 Identification of a gene coding for a new squamous cell carc... | ACCEPT | Summary: DSE has a supported Golgi-associated membrane pool. Reason: The original localization study used human full-length SART2/DSE fused C-terminally to GFP. COS7 microscopy and sucrose-gradient fractionation supported predominantly ER and partial Golgi localization; MCF-7 transfectants showed a similar organelle pattern. Retain the curated membrane assignment, while distinguishing tagged-protein localization from native untagged occupancy and from experimentally mapped membrane topology. The paper explicitly leaves native localization unresolved. |
| GO:0000139 Golgi membrane | IEA GO_REF:0000044 | ACCEPT | Summary: The UniProt Golgi-membrane mapping is supported. Reason: The exact SL-0134 source location is present in the official DSE record and is grounded in PMID:10679095. Tagged human DSE localization and the independent Reactome catalyst/entity placement support the Golgi membrane assertion. This does not establish an exclusive Golgi pool or a membrane-spanning topology. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0134 SUPPORTS TRANSFER The literal source location is present; original tagged-protein and curated reaction-compartment evidence agree. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-2022052 | ACCEPT | Summary: The Reactome catalyst is explicitly localized to the Golgi membrane. Reason: The reaction R-HSA-2022052 links catalyst activity 2022049 to the DSE/DSEL set R-HSA-3734047 and its DSE member R-HSA-2022047, which explicitly identifies Q9UL01 residues 23–958 at the Golgi membrane. This is target-resolved curated localization, corroborated by the tagged-protein study; the set is an alternative-enzyme set, not a required DSE–DSEL complex. |
| GO:0005783 endoplasmic reticulum | IDA PMID:10679095 Identification of a gene coding for a new squamous cell carc... | MODIFY | Summary: Specify the supported ER membrane association. Reason: The original tagged human DSE construct has a predominant ER-like microscopy pattern and cofractionates with an ER membrane marker. Together with the independently curated ER-membrane assertion, this supports the more specific ER membrane term. Human-fibroblast microsomal epimerase activity corroborates membrane association but does not distinguish ER from Golgi. This location refinement does not assign the epimerization reaction to the ER, claim experimentally resolved topology or exact native distribution, or exclude Golgi and vesicle pools. Proposed replacements: endoplasmic reticulum membrane |
| GO:0005789 endoplasmic reticulum membrane | EXP PMID:10679095 Identification of a gene coding for a new squamous cell carc... | KEEP AS NON CORE | Summary: Retain the positive ER-membrane pool as non-core localization. Reason: The original study reports predominant ER-pattern localization and ER-marker cofractionation for human full-length DSE/GFP in COS7 cells, with a similar ER/Golgi pattern in human MCF-7 transfectants. This positive ER pool is retained without treating it as false, transient or background. The study does not measure ER-localized epimerization and explicitly leaves native untagged localization unresolved. Human-fibroblast microsomal activity supports membrane association but does not resolve ER from Golgi. The compact catalytic core uses the target-resolved Golgi reaction compartment; ER occupancy remains a supported non-core location, not a rejected observation. Supporting Evidence: PMID:10679095 This gene product had a characteristic of an endoplasmic reticulum-resident protein. |
| GO:0005789 endoplasmic reticulum membrane | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Retain the supported ER-membrane source mapping as non-core. Reason: The official DSE record contains SL-0097 with the original positive localization evidence. Tagged human DSE establishes an ER-associated pool, while human-fibroblast microsomal activity independently supports membrane association without resolving individual organelles. The electronic and experimental ER-membrane assertions are retained at the same non-core scope: neither is false, and neither alone assigns catalytic epimerization to the ER. The core identifies the Golgi reaction compartment without claiming exclusive protein occupancy or an inferred membrane topology. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: UniProtKB-SubCell:SL-0097 SUPPORTS TRANSFER The exact curated source location is present and independently supported by bounded localization and membrane-fraction evidence. |
| GO:0005794 Golgi apparatus | IDA PMID:10679095 Identification of a gene coding for a new squamous cell carc... | MODIFY | Summary: Specify Golgi membrane rather than the whole Golgi apparatus. Reason: Tagged human DSE shows a partial Golgi localization pattern and cofractionation with the study’s Golgi marker. The original EXP membrane assertion and target-resolved Reactome Golgi-membrane catalyst placement support this more informative compartment. The assay used tagged constructs and does not resolve native topology or eliminate the larger ER pool. Proposed replacements: Golgi membrane |
| GO:0015012 heparan sulfate proteoglycan biosynthetic process | IEA GO_REF:0000041 | REMOVE | Summary: The heparan-sulfate pathway assignment conflates distinct epimerase systems. Reason: The imported UniPathway mapping is present in DSE’s current UniProt pathway annotations, but it conflicts with substrate-specific biology. PMID:7092807 experimentally distinguishes dermatan/chondroitin and heparan epimerase activities by substrate, pH and cofactor conditions. Human DSE catalyzes chondroitin-to-dermatan epimerization (EC 5.1.3.19; RHEA:21084), whereas GLCE catalyzes epimerization of N-sulfated heparosan (EC 5.1.3.17; RHEA:20197). The corresponding official EC entries explicitly distinguish these reactions. This is a source-pathway assignment error, not rejection based on missing full text or on a generic family name. Propagation Review Root cause: SOURCE BAD Sources checked: UniPathway:UPA00756 SOURCE BAD The heparan-sulfate pathway itself is valid, but its assignment to DSE is contradicted by the distinct demonstrated substrate/reaction scope. This classification concerns the target-source mapping, not pathway database quality. Supporting Evidence: PMID:7092807 two different uronosyl epimerases, one for the biosynthesis of heparan sulphate and one for that of dermatan sulphate. |
| GO:0030659 cytoplasmic vesicle membrane | EXP PMID:10679095 Identification of a gene coding for a new squamous cell carc... | KEEP AS NON CORE | Summary: Retain the contextual vesicle-membrane localization. Reason: The original paper reports a secretory-vesicle pattern for full-length human DSE/GFP in COS7 cells; that pattern was not seen in the human MCF-7 comparison. It is a positive, host/construct-dependent localization observation. Retain the curated membrane annotation without assigning catalytic work in vesicles, constitutive native occupancy, or secretion of DSE itself. The biosynthetic core is associated with the Golgi membrane. |
| GO:0030659 cytoplasmic vesicle membrane | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Retain the electronic vesicle-membrane mapping at the same contextual scope. Reason: SL-0089 is explicitly present in the official DSE record and traces to the tagged-protein localization study. Its positive vesicle observation is retained with the same COS7 versus MCF-7 limitation as the experimental assertion. This is contextual location evidence, not a separate enzymatic function or universal endogenous localization. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: UniProtKB-SubCell:SL-0089 SUPPORTS TRANSFER The source location is present and the original observation is positive, but its construct/host-dependent scope is contextual. |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | EXP PMID:19004833 Identification of the active site of DS-epimerase 1 and requ... | ACCEPT | Summary: Human DSE has experimentally established chondroitin epimerase activity. Reason: Sequence-verified human NM_001080976 constructs and isolated tagged wild-type/mutant proteins were tested for epimerase activity in the original study. The measured net GlcA-to-IdoA reaction supports this molecular function. The 2009 proposed catalytic mechanism was subsequently revised by structural work; that revision does not negate the measured activity and does not establish glycan-degrading lyase activity. Supporting Evidence: PMID:19004833 proper N-glycosylation of DS-epimerase 1 is required for enzyme activity. |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | EXP PMID:7092807 Biosynthesis of dermatan sulphate. Assay and properties of t... | ACCEPT | Summary: The historical human-microsome assay supports the epimerase reaction. Reason: The original Methods explicitly prepare enzyme from cultured human skin fibroblasts and assay C5 tritium release from labelled chondroitin. Membrane fractionation and substrate/cofactor comparisons establish the dermatan-forming activity, but the study predates gene identification and does not isolate molecularly identified DSE from DSEL. Later human recombinant DSE work supplies the gene-level link, so the curator’s correct activity assertion is retained. |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | IBA GO_REF:0000033 | ACCEPT | Summary: The PAINT epimerase assignment is consistent with the recorded ancestral node and direct target evidence. Reason: The literal IBD for GO:0047757 is at PTN001036588, upstream of the human DSE leaf PTN002496188 in PTHR15532. The ungapped target alignment exactly matches the official 958-residue sequence. Its descendant evidence includes human DSE, bovine DSE/P0C2H4 and mouse Dsel/Q0VBN2 (MGI:2442948); the latter is correctly distinguished from mouse Dse. Direct target enzymology corroborates the inherited activity. Target self-inclusion is expected experimental grounding, not circularity. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN001036588 SUPPORTS TRANSFER The literal IBD, 17-node target path and exact target MSA match are verified. Bovine and human activity plus mouse Dsel perturbation evidence support the family-level placement; no target-specific loss is established. |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | IDA PMID:16505484 Biosynthesis of dermatan sulfate: chondroitin-glucuronate C5... | ACCEPT | Summary: Recombinant human DSE directly catalyzes the epimerization reaction. Reason: The original study identifies the bovine spleen enzyme and separately tests sequence-verified full-length human SART2/DSE cDNA with a C-terminal His tag in mammalian expression systems. Tritium-release assays and carbon-14 product mapping support GlcA-to-IdoA conversion on chondroitin. The human construct evidence is distinct from bovine discovery and does not rely solely on a disease phenotype. Supporting Evidence: |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | IEA GO_REF:0000120 | ACCEPT | Summary: The exact Rhea/EC mapping matches DSE chemistry. Reason: RHEA:21084 records chondroitin 4′-sulfate to dermatan 4′-sulfate and links EC 5.1.3.19; the official EC definition matches the GO reaction. DSE also epimerizes unsulfated chondroitin in the primary experiments, so the registry equation is not treated as an exclusive sulfation requirement. This mapping is distinct from GLCE’s heparan-substrate reaction. Propagation Review Root cause: NO FAILURE CORE Sources checked: RHEA:21084 SUPPORTS TRANSFER The actual Rhea equation, linked EC identifier, GO definition and independently measured DSE reaction agree. |
| GO:0047757 chondroitin-glucuronate 5-epimerase activity | TAS Reactome:R-HSA-2022052 | ACCEPT | Summary: The Reactome catalyst identity and activity match DSE. Reason: R-HSA-2022052 links catalyst 2022049, explicitly labelled GO:0047757, through the DSE/DSEL alternative-enzyme set to Q9UL01 in R-HSA-2022047. Human biochemical evidence corroborates the reaction. Set membership does not imply an obligate heteromer or confer DSEL-specific properties on DSE. |
| GO:0050651 dermatan sulfate proteoglycan biosynthetic process | IDA PMID:16505484 Biosynthesis of dermatan sulfate: chondroitin-glucuronate C5... | ACCEPT | Summary: DSE performs a chemical step of dermatan-sulfate proteoglycan biosynthesis. Reason: The enzyme converts uronic-acid stereochemistry within chondroitin chains. The original human-expression experiments additionally measured increased iduronic acid and iduronate blocks in metabolically labelled cellular CS/DS chains. DSE therefore performs biosynthetic work itself; this is not inferred merely from necessity, substrate consumption or a downstream disease phenotype. |
| GO:0050651 dermatan sulfate proteoglycan biosynthetic process | IEA GO_REF:0000117 | ACCEPT | Summary: The ARBA biosynthesis assignment matches the target’s demonstrated chemistry. Reason: ARBA00095326 requires FunFam 1.50.10.100:FF:000001 and 2.70.98.70:FF:000001 in Primates. Both exact FunFam memberships and the target’s primate lineage are present in the official DSE record. The predicted process is independently established by human DSE-mediated conversion of cellular CS/DS chains, rather than inferred from a generic glycosylated-protein feature. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00095326 SUPPORTS TRANSFER The actual rule conditions match the captured target record, and primary human chemistry supports the process. |
| GO:0050651 dermatan sulfate proteoglycan biosynthetic process | TAS Reactome:R-HSA-2022923 | ACCEPT | Summary: DSE is an explicit enzymatic participant in the Reactome dermatan-sulfate pathway. Reason: R-HSA-2022923 contains reaction R-HSA-2022052, whose catalyst chain resolves to human DSE/Q9UL01 at the Golgi membrane. DSE performs the GlcA-to-IdoA conversion, rather than merely being necessary for the pathway or supplying an unmodified substrate. The assignment does not transfer the pathway’s separate sulfotransferase or secretion reactions to DSE. |
| GO:0050654 chondroitin sulfate proteoglycan metabolic process | IBA GO_REF:0000033 | ACCEPT | Summary: Chondroitin-chain modification supports the inherited metabolic-process assertion. Reason: The literal GO:0050654 IBD is at PTN001036588 and traces to mouse Dsel (MGI:2442948/Q0VBN2), whose current IMP evidence cites PMID:22496542. Its verified abstract reports altered CS/DS composition after Dsel deletion; the full original was not recovered. Independently, human DSE directly modifies glucuronic-acid residues in chondroitin chains. Thus the target performs chemical work in chondroitin proteoglycan metabolism as well as dermatan biosynthesis. The broad term is not reinterpreted as chain initiation, polymerization, degradation or a mouse brain phenotype. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN001036588 SUPPORTS TRANSFER The exact IBD and target descent are verified. The mouse Dsel source has current experimental grounding, and direct human chondroitin epimerization supports the inherited process; donor count is not a weakness. Supporting Evidence: PMID:22496542 a 38% reduction in iduronic acid compared with wild-type littermates |
| GO:0050655 dermatan sulfate proteoglycan metabolic process | IBA GO_REF:0000033 | MODIFY | Summary: Specify the established biosynthetic branch of dermatan-sulfate metabolism. Reason: The literal GO:0050655 IBD at PTN001036588 is supported by mouse Dsel and direct human DSE evidence. Human DSE performs the epimerization that creates iduronic-acid-containing dermatan regions, with cellular product analysis corroborating biosynthesis. GO:0050651 is a verified descendant of the imported metabolic term and captures this demonstrated direction. This does not imply that DSE degrades dermatan sulfate or that the original ancestral placement was incorrect. Propagation Review Root cause: TERM SCOPING PROBLEM Sources checked: PANTHER:PTN001036588 SUPPORTS TRANSFER The source and inheritance are supported; the target’s directly demonstrated biosynthetic role warrants the narrower descendant. Proposed replacements: dermatan sulfate proteoglycan biosynthetic process |
Loading supporting content…
Download this section (compressed HTML)Q: How is endogenous DSE distributed between ER, Golgi and vesicular membranes, and which pools contribute measurably to dermatan-sulfate chain epimerization?
Q: How does a defined chondroitin chain bind the catalytic domain, and which proposed proton-transfer steps explain the measured net epimerization under native membrane conditions?
Experiment: Combine endogenous DSE tagging with genetic specificity controls, quantitative organelle fractionation and chain-composition analysis to distinguish native compartment occupancy from tagged-protein trafficking.
Experiment: Test sequence-defined DSE catalytic variants on chemically defined chondroitin substrates and pair product stereochemical analysis with substrate-bound structural work, distinguishing epimerization from any transient or net bond cleavage.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)