The review distinguishes actual original evidence from database inference and from successful download status. Normal publication-cache bytes and their availability metadata were not changed when originals were recovered separately. The reading limits below apply to the reviewed experiments, including uninspected figures and supplements.
| Reference and original routes | Operative read scope and limits |
|---|---|
| PMID:7651464 — Original | Positive abstract describes human XPG in insect cells, nuclear IF and substrate preferences. Full original unavailable. Independent human nuclear-foci evidence supports localization without assuming the original kinetics or isoform behavior. |
| PMID:7657672 — Original | Positive abstract only; original JBC assays unavailable. Independent human reconstitution and catalytic-mutant experiments support the broad NER role, without claiming inspection of this paper’s antibody experiments. |
| PMID:7700386 — Original | Nature preview and positive abstract only. The original RPA-associated claim remains undecided; it is not restated as constitutive RPA-subunit membership. |
| PMID:8078765 — Original; Additional route | Original scanned pp. 3312–3315 and Figures 1–4 inspected. Complete human coding frame in Sf9, purified XPG, circular ssM13 incision, Mg dependence and termini assays. Weak dsDNA nicking is not UV-preferential. Page 3316 continuation unread; no isolated-ssDNA affinity claim. |
| PMID:8090225 — Original | Positive abstract and Nature preview only. Original Methods/Results/figures unavailable. Independent human incision and NER reconstitution resolve the broad catalytic and repair claims; exact original assays are not presented as read. |
| PMID:8247134 — Original | Verified yeast Rad2 publisher abstract only; apparent PDF route returned abstract HTML. Original Methods/Results/figures unavailable. Independent human PMID:8078765, rather than cleavage inferred from junction assays, resolves target ssDNA endonuclease activity. |
| PMID:8652557 — Original | Positive abstract only; ACS original unavailable. Human XPG–XPD/XPB support comes from independently read PMID:31253769. Exact p62, p44 and CSB pairs remain unresolved. |
| PMID:8710877 — Original; Additional route | Positive UV-redistribution abstract only; original operative figures unavailable. Independent human nuclear localization and UV-complementation studies resolve the broad claims. No direct read of the 1996 redistribution experiment is claimed. |
| PMID:8855246 — Original; Additional route | Original scanned Methods/Results pp. 10718–10721 and Figure 3 image inspected. Yeast Rad2–TFIIH purification/co-IP and functional NER reconstitution. Page 10722 unread. This is yeast donor evidence, separately corroborated by human assembly experiments. |
| PMID:9305916 — Original; Additional route | Author original Methods and Results Figures 4–6 read: human XPG 981–1009 fragment and full-length insect-expressed human XPG, PCNA beads, competition and DNase controls. No pixels inspected; coupling to repair synthesis remains a mechanistic interpretation. |
| PMID:9927729 — Original; Additional route; Additional route | Publisher original Methods/Results/Discussion and all three figure images inspected. Reciprocal association, NTHL1 substrate binding and combined glycosylase/AP-lyase product assay support activation. The protein-preparation caveat remains; the discussion model for transcription-coupled BER is not adopted. |
| PMID:10026181 — Original; Additional route | Author original operative Methods/Results Figures 4–7 read: human WT/catalytic-mutant repair and engineered WT/ΔN136 self-association, with mixing controls and high-salt monomer measurements. No figure pixels inspected; ΔN136 is not a natural splice product. |
| PMID:11259578 — Original | Original main Methods, Results Figures 1–8 and Discussion read. Native human factors, CDK7 capture and cisplatin dual-incision assays distinguish functional assemblies from a permanent repairosome. No figure pixels, full tables or supplement inspected. |
| PMID:12644470 — Original; Additional route | Author original Methods and Results Figures 1–4, 6 and 8 read. Human His-tagged WT/E791A; junction cleavage, competition and footprinting. Figure 4 locates short ss-overhang contacts as well as duplex contacts; isolated strands compete poorly. No figure pixels inspected. |
| PMID:15358233 — Original | Verified abstract explicitly includes human FLAG-XPG despite the title foregrounding other proteins. Main assays unavailable. Independent PMID:9927729 supports NTHL1 association and activation; no wrong-target inference. |
| PMID:16167068 — Original; Additional route | Author original Methods, Results Figures 1–5 and Discussion read: primary human fibroblast UVC dose/time responses and WT/E791A complementation. Lentiviral selection and mutant limits retained. No figure pixels inspected. |
| PMID:16246722 — Original; Additional route | Positive abstract only; operative main experiments unavailable. Independent bubble-binding and human catalytic-pathway evidence support those broad claims, but RNA polymerase II/CSB interaction and upstream recognition assays remain unverified. |
| PMID:17208056 — Original; Additional route | Partial publisher Methods/Results and positive abstract only: human fibroblasts, 254-nm exposure and initial CD95 section. Complete downstream assays and figures unavailable. Independent PMID:16167068 supports UV-C response, not every source-specific mechanism. |
| PMID:17268471 — Original; Additional route | Main article unavailable. Actual Supplementary Figure S5 caption/experiment and relevant Methods read: human XPG–FLAG-GADD45A co-IP in HEK293T nuclear extracts with IgG and FLAG-Smad1 controls. No pixels inspected or XPG demethylase activity inferred. |
| PMID:26833090 — Original | Original Results Figures 4–7 and operative Methods read: controlled human nuclear foci, HRR assemblies, DR-GFP repair and repair-factor loading. P100 includes chromatin, nuclear matrix and other insoluble material. No figure pixels or exact supplementary construct coverage claimed. |
| PMID:31253769 — Original; Additional route | Original main Methods, operative Results and captions read across the repair-assembly assessment: full-length human XPG, Hi5 expression/tag removal, motor-stimulation assays and BS3 crosslinking. XPG is absent from final cryo-EM density; E791A is used in that preparation. No figure pixels or detailed Supplementary Data 1 parsing. |
| PMID:32522879 — Original; Additional route | Original main text and supplementary Online Methods pp. 2–5 read; main Figure 3 and Figure 4 pixels inspected. Full-length human XPG is separate from the archaeal-segment XPGcat hybrid. Bubble/Y cleavage, coexpression-dependent association, SAXS and EM have construct-specific scope. |
| PMID:32821917 — Original | Original Methods, Results and Figure 5 captions read. Human residues 1–112 linked to 750–990; synthetic splayed-arm DNA, mutant incision and binding assays. No natural-isoform equivalence, damaged-substrate claim or intact NER structure inferred. No figure pixels or supplement inspected. |
| PMID:35512704 — Original | Original main Methods and relevant Results read, including BRET screening and mutant/WT comparisons. Exact ERCC5–BRAF supplementary spreadsheet record unavailable. Selected BRAF–KEAP1 follow-up is not transferred to XPG. |
The actual PANTHER PTHR16171 tree places human PTN002497156 below the Eukaryota assertion node PTN000792422. The ungapped target alignment exactly matches the 1,186-aa P28715 sequence. Three IBD assertions were traced: GO:0006289, GO:0000112 and GO:0000014, with no target-path NOT. A single yeast donor is legitimate phylogenetic support; human self-WITH is legitimate descendant evidence, not circularity. Yeast donor experiments and independent human experiments retain their separate species and preparation scopes.
The official UniProt target and InterPro matches verify IPR001044, IPR006085, IPR006086, IPR008918, IPR019974 and IPR029060. Broad family/domain mappings are evaluated against the human experiments, not treated as new experimental evidence.
| Rule | Verified condition set, zero-based | Biological treatment |
|---|---|---|
| ARBA00029205 | Set 2: IPR006085, IPR029060 and Primates | DNA endonuclease supported by human incision |
| ARBA00086510 | Set 2: FunFams 1.10.150.20:FF:000037, 3.40.50.1010:FF:000022 and 3.40.50.1010:FF:000023 | NER supported by target incision |
| ARBA00028026 | Set 11: IPR006085, IPR029060 and Primates | Refine recombination to experimentally supported homologous repair |
| ARBA00028103 | Set 5: the same three FunFams | Broad UV response retained as non-core |
All 11 source Reactome roots have actual forward routes to human entity 67447, reference entity 67448, P28715, modeled residues 1–1186 in nucleoplasm. Location roots are 5689317, 5690213, 5690988, 5690990, 5690991, 6782138, 6782141, 6782204, 6782211 and 6782224. The separate pathway root is 6781827. Neighboring polymerase recruitment or ERCC1–ERCC4 incision does not confer those activities on XPG.
Two catalysts identify XPG itself: GG-NER event 5690988 → catalyst 5691069 → active unit 67447; and TC-NER event 6782224 → catalyst 6782216 → active unit 67447. Both have DNA endonuclease activity. TC-NER pathway 6781827 contains dual-incision branch 6782135 and the XPG incision event, establishing an executed step. This does not independently verify every upstream recognition or backtracking model or imply access to PMID:16246722.
The official HPA ERCC5 record reports supported nucleoplasm localization: HPA045845 in U-251MG and HPA050374 in A-431, U-251MG and U2OS. It explicitly cautions that the antibodies can recognize proteins from multiple genes. HPA image pixels were not inspected. Controlled human XPG nuclear-foci experiments and explicit Reactome target compartments independently corroborate the location. The P100 fraction in PMID:26833090 is not purified chromatin.
Current QuickGO definitions were checked for the authored refinements: bubble DNA binding, DNA endonuclease activity, homologous double-strand-break repair, nucleoplasm, chromatin, enzyme activator activity, TFIIH-class factor binding and TFIIH holo-complex binding. DNA endonuclease is not currently a formal descendant of GO:0016788 in the captured graph; that refinement specifies the demonstrated chemistry without inventing an ontology edge. The factor-3 complex definition has a yeast example but is not defined as yeast-only.
| Natural product | Sequence reconstruction | Interpretation |
|---|---|---|
| P28715-1 | 1,186 aa, full-length canonical sequence | Distinguish each experimental preparation and tag |
| P28715-2 | 419 aa; VSP_035380 removes residues 1–767 | No blanket transfer of full-length activities or location |
| P28715-3 | 232 aa; VSP_053828 replaces residues 225–232 and VSP_053829 removes residues 233–1186 | No equivalence to engineered catalytic cores or deletion mutants |
The source sequence edits were replayed against the official UniProt sequence. These metadata establish product identities, not the products’ physiological functions. No new process annotation is proposed to fill an assumed pathway gap.
The core uses current DNA endonuclease activity, nucleotide-excision repair, nucleoplasm and nucleotide-excision repair complex. The complex term covers NER-protein assemblies; it does not require a permanent preassembled complex. Human incision assemblies 5690214 and 6782219 each explicitly contain ERCC5 entity 67447, the active unit of their respective DNA-endonuclease catalysts. PMID:11259578 Methods and Results Figures 3–6 supply independent functional human complex evidence with ionic-strength limits. The core preserves the distinct natural products and keeps HR recruitment and NTHL1 activation separate from nuclease chemistry.
The full binding refinement retains bubble DNA binding together with double-stranded and single-stranded DNA binding. PMID:12644470 Figure 4 resolves duplex and short single-stranded junction contacts; Figure 2 competition distinguishes these from unrestricted isolated-strand binding. The current official is-a/part-of closures show that TFIIH-class factor binding and TFIIH holo-complex binding are not ancestors or descendants of each other. Their separate subunit-versus-holo-complex evidence is retained.
PMID:11259578 Figure 6 Results explicitly retain functionally active XPG with TFIIH after 150 mM KCl washing and lose that interaction at 300 mM KCl. This is evidence for a condition-dependent repair subassembly, compatible with the absence of a permanent all-factor repairosome. The original Methods and Results were reread for this distinction; no new figure-pixel inspection is claimed. PMID:15358233 remains abstract-only but explicitly reports FLAG-XPG-dependent stimulation of NTH1 glycosylase/AP-lyase activity. The unread supplement and figure pixels for PMID:32821917 remain the boundary on adjudicating its damaged-DNA binding assertion.