title: DSE source and assay guide
gene: DSE
annotation_scope:
imported_annotations: 21
new_annotations: 0
All 21 raw annotation contexts produce 21 preserved source assertions. The review identifies one direct epimerase mechanism, distinguishes dermatan/chondroitin from heparan substrates, and retains the source-specific limits of tagged localization and phylogenetic inference.
| Source | Original location and inspected scope |
|---|---|
| PMID:7092807(https://pmc.ncbi.nlm.nih.gov/articles/PMC1163673/) | All original scanned pages 489–493 were visually inspected. Page 490 Methods specifies cultured human skin fibroblasts as the enzyme source; pages 491–493 contain the activity and substrate/cofactor comparisons. Table 1 on page 492 resolves membrane fractions, not individual organelles. Figure 4 and the surrounding Results compare the dermatan and heparan systems. The normal cache is abstract-only. |
| PMID:10679095(https://pubmed.ncbi.nlm.nih.gov/10679095/) — indexed author-uploaded original | Original cloning/construct Methods, microscopy and fractionation Methods, localization Results, Figure 5 caption and Discussion windows were inspected. Human AF098066/SART2-GFP and terminal deletions were expressed in COS7 and MCF-7 cells. Predominant ER and partial Golgi localization are positive; the vesicle pattern differs by host. Native untagged localization and membrane topology remain unresolved. No figure-pixel or complete original-PDF read is claimed. A publisher PDF route returned a client challenge, not a paper. |
| PMID:16505484(https://pubmed.ncbi.nlm.nih.gov/16505484/) — Lund author manuscript | Operative original Methods and Results were independently read. Bovine spleen discovery is separate from sequence-verified human IMAGE5272885 full-length cDNA with a C-terminal His tag. Tritium-release assays and carbon-14 product mapping establish GlcA-to-IdoA conversion; cellular labeling establishes altered CS/DS composition. No substrate depolymerization is inferred. The normal cache remains abstract-only. |
| PMID:19004833(https://pubmed.ncbi.nlm.nih.gov/19004833/) — indexed author original | Original Methods and Results were independently read. Human NM_001080976 tagged wild-type and mutant proteins were isolated from mammalian expression systems and assayed for epimerase activity. Mutagenesis and glycosylation findings are experimental; the proposed His450-base/cleavage-religation mechanism is a homology-based model subsequently revised. The normal cache remains abstract-only. |
| PMID:33815739(https://pmc.ncbi.nlm.nih.gov/articles/PMC8006597/) — full XML | The genuine full original was independently read. Human 23–775 protein yielded a 2.41 Å crystal; 23–894 is a separate soluble experimental construct. H452/E470/N481 coordinate the metal, and the revised catalytic proposal involves H205/Y261/Y473. No substrate cocrystal or measured net glycan-degrading lyase activity is claimed. The normal additional cache contains full XML. |
| PMID:22496542(https://pubmed.ncbi.nlm.nih.gov/22496542/) — institutional abstract | Complete abstract and current mouse donor records were inspected. Mouse Dsel loss alters brain CS/DS iduronic acid and sulfation. Publisher full-text access failed; no detailed Methods or supplementary read is claimed. The normal additional cache is abstract-only. |
The official Q9UL01 sequence record and FASTA agree on 958 residues and contain no alternative-products block. The predicted signal peptide and annotated processed chain do not establish separate functions. Structural constructs remove native sequence and are not natural isoforms. These distinctions apply to all activity and localization conclusions.
GO:0047757, EC 5.1.3.19 and RHEA:21084 identify DSE's chondroitin-to-dermatan epimerization. The actual Rhea TSV query links the chondroitin 4′-sulfate/dermatan 4′-sulfate equation to that EC identifier. Primary assays using unsulfated chondroitin show that the registry equation is not an exclusive substrate-sulfation rule.
EC 5.1.3.17 and GLCE/O94923 instead concern N-sulfated heparosan epimerization in heparan/heparin biosynthesis. The original 1982 substrate/pH/cofactor comparisons distinguish the two enzyme systems. The source UPA00756 heparan assignment is actually present in DSE's current UniProt pathway annotation, but this target mapping conflicts with the reaction evidence. The electronic annotation is rejected on that biological ground. The standalone UniPathway endpoint and mapping download were unavailable; no claim is made to have read their absent responses. No experimental annotation is removed.
GO:0050651 dermatan-sulfate proteoglycan biosynthesis is a descendant of GO:0050655 dermatan-sulfate metabolism. The corresponding IBA refinement adds supported specificity. GO:0050654 chondroitin-sulfate proteoglycan metabolism remains biologically appropriate because DSE modifies the chain directly; neither this term nor the dermatan biosynthetic term is an ancestor of the other in the inspected is_a/part_of closures. Both are included in the same catalytic core because DSE directly performs the chain modification, without asserting polymerization or degradation.
GO:0005789 ER membrane and GO:0000139 Golgi membrane refine their respective organelle terms. The refinement combines original tagged-protein microscopy/fractionation, existing curated membrane assertions and reaction-compartment evidence. It does not claim exclusive occupancy or measured membrane topology. The COS7 vesicle observation is retained at its host/construct-dependent scope. The HPA DSE XML has no subcellular assay section in the inspected record; tissue expression is not substituted for localization.
The two positive ER-membrane assertions are retained as non-core, and the broader ER assertion is still refined to ER membrane. The tagged-human localization experiment supports that pool but does not measure its catalytic contribution. The human-fibroblast microsomal activity does not resolve ER from Golgi. The core follows the target-resolved Golgi reaction compartment; this does not deny the ER pool or label it transient or background. Location specificity and assignment of catalytic work to a compartment are distinct judgments.
The exact public Reactome chain is R-HSA-2022923 pathway → R-HSA-2022052 reaction → 2022049 catalyst → R-HSA-3734047 DSE/DSEL set → R-HSA-2022047 DSE entity. The catalyst specifies GO:0047757; the member identifies Q9UL01 residues 23–958 at the Golgi membrane. The corresponding ContentService records were inspected. Set membership denotes alternative enzymes, not an obligatory heteromer. No target activity was found in the inspected GO-CAM index, and no annotation is added on the basis of absence.
The PANTHER PTHR15532 family tree contains 95 nodes. Human DSE leaf PTN002496188 has a 17-node path through PTN001036588; its ungapped alignment exactly matches the 958-residue official sequence. Three literal PAINT IBD records at that node cover GO:0047757, GO:0050655 and GO:0050654. This is an ancestral placement, not a pairwise similarity transfer.
The source identities are human Q9UL01, bovine P0C2H4, and mouse Dsel Q0VBN2/MGI:2442948. Mouse Dsel is distinct from mouse Dse/Q8BLI4. Its current QuickGO annotations supply IMP evidence from PMID:22496542 for the three shared functions. Human recombinant activity and bovine discovery are supported by PMID:16505484. The common ancestral judgment is compatible with these paralog-specific observations; target self-inclusion is legitimate experimental grounding.
ARBA00095326 requires FunFam 1.50.10.100:FF:000001, FunFam 2.70.98.70:FF:000001 and Primates/NCBITaxon:9443. All three conditions match the official DSE record. The independent biochemical evidence supports the emitted dermatan-sulfate biosynthesis term; a matching rule is not treated as independent experimental evidence.
The normal caches for PMID:7092807, PMID:10679095, PMID:16505484 and PMID:19004833 remain abstract-only, even where separately recovered originals were inspected. PMID:33815739 has a genuine full XML cache; PMID:22496542 is abstract-only. The pre-existing sequence-resource cache PMID:15489334 is preserved without a new functional or full-original-read claim. The two pre-existing Reactome caches and GO-CAM index are unchanged. Failed routes and client challenges are retained as access provenance, not biological evidence.
The manual reference-level availability flag describes original access rather than the contents of the normal cache. It is false for the recovered scans/manuscript/XML and for accessible indexed original text in the documented sections. For PMID:10679095 and PMID:19004833, this does not claim a complete original PDF or unexamined supplement. It is true for PMID:22496542, whose authentic abstract still supports the limited finding for which it is cited. Citation correctness and full-text availability are separate assessments.
The original-only quotation locators are the localization Results immediately following Figure 5 in the indexed PMID:10679095 author version; Experimental Procedures, plasmid construction, in the indexed PMID:19004833 author version; and the expression Methods in the Lund PMID:16505484 manuscript. These short fragments are stored in supporting_text_fulltext, not represented as cache substrings. Each cached structural fragment from PMID:33815739 occupies a separate support entry and is an exact contiguous substring. Its metal-site substitutions either abolished activity or prevented protein expression; they are not described as uniform catalytic loss in successfully expressed proteins. The catalytic model remains distinct from a substrate cocrystal.
The 21 source dictionaries, one official sequence boundary and all machine source bytes are preserved. Thirteen assertions are accepted, three refined, four retained as non-core locations and one electronic pathway assertion rejected; there are no new annotations. Genuine configured research-provider attempts failed and yielded no report. Aggregate quoted words across the YAML and notes, including the inherited excerpt, are 20 for PMID:16505484, 22 for PMID:10679095, 20 for PMID:19004833, 18 for PMID:7092807, 19 for PMID:33815739 and 10 for PMID:22496542. Each excerpt is used once, and this guide adds no quotations.