In-progress assessment, 2026-09-28. The 50 original annotations and both alternative products remain unchanged. All 17 original PMID abstracts and four complete cached Reactome summaries have been read; this does not establish every source-specific interaction. The configured research provider and fallback failed once without producing a report. Four additional normal publication fetches each completed with a DNS failure and no output; their separate receipts preserve the actual attempts.
PMID:18234692 identity/abstract verified against PubMed. The primary author manuscript is available at https://eprints.lib.hokudai.ac.jp/repo/huscap/all/38124/STAT3-II.pdf. Root read the complete Methods and Results text, relevant Discussion, and the introduction; figure pixels and supplementary material have not been audited. The cache remains abstract-only and must not be rewritten. Human Hep3B, HeLa and 293T experiments support STAT3 binding, nuclear retention and associated transcriptional output. A zebrafish BART construct rescues the human knockdown phenotype. Reduced phospho-STAT3 is rescued by vanadate; leptomycin B restores nuclear accumulation. These findings distinguish retention from an intrinsic kinase or direct DNA-binding activity. ARL2 expression enhances BART–STAT3 association, and its nucleotide-state mutants affect association and STAT3 accumulation. Do not infer that all overexpression constructs are human merely from the host cell line.
GO:0030695 is defined by GO as binding and modulation of GTPase activity; it is not restricted to GAP activity. The 2021 paper is therefore essential before judging the existing regulator assertion. GO:0051020 is GTPase binding; its small-GTPase child should be checked if proposed. No OLS MCP is available in this session; primary GO/QuickGO pages are used instead. No final action table, new process annotation or core function has been authored. Existing IBA self-inclusion is legitimate and must not be called circular. Unverified screen-specific interactions will remain explicitly unresolved without asserting absence or misattribution.
2026-09-28: root read the complete abstract and targeted original 2013 report text in the author laboratory's copy, including the HYDIN differential diagnosis, mouse retinal localization, human ARPE19 depletion, SK-N-SH interaction/construct experiments, nasal epithelial localization, and the concluding interpretation. The reading included Figures 3–4 textual captions and embedded methods, but not supplementary figures/tables, image pixels or an independent clinical-genetic reanalysis.
Human ARPE19 and nasal epithelial observations corroborate basal-body localization. ARL2 depletion changes that localization, whereas ARL3 depletion did not in those tested cells. The 2013 authors' broader ARL3-independent interpretation concerns this localization experiment; it does not negate the distinct 2021 purified-protein evidence for cooperative ARL3 exchange. The relation between these biochemical and cellular observations remains a question for tissue-specific experiments. The human ARL2BP/bovine ARL2 construct distinction and the separate HYDIN explanation for the second proband's respiratory/hearing phenotype are explicit in the primary report.
The four additional references now have normal publication records with verified identities. PMID:23849777 remains abstract-only in the cache despite the earlier external primary-PDF access. PMID:29718757, PMID:31425546 and PMID:33438581 contain XML-extracted full text. Availability flags describe these immutable records, separately from what was actually read.
The 2021 co-GEF paper's main Results, Discussion and Methods were read in the normal cache, including GDP-to-GDP exchange controls, effector-mediated BART turnover, ARL2/ARL3 GTP-state stabilization, construct purification and live imaging in human RPE and mouse NIH/3T3 cells. Its model does not establish a stable BART–ARL13B complex, nor does ciliary colocalization directly measure the reaction in a photoreceptor. Figure captions were read; image pixels and underlying numerical source data were not reanalysed. PMID:33438581(https://doi.org/10.7554/eLife.64624)
For the 2018 mouse study, the main Results through doublet ultrastructure, early Discussion through alternative mechanisms, and mouse-generation, antibody and localization Methods were read. Knockout controls support the localization experiments. The connecting-cilium length is preserved while distal axonemes shorten and B-tubule inner junctions remain open. Proposed inner-junction membership, microtubule association and transport of structural factors remain alternatives. Attempts to recover ARL2 association from photoreceptors were unsuccessful; this limits the tissue mechanism without invalidating purified human binding. The complete remaining Methods, raw images and supplements were not audited. PMID:29718757(https://doi.org/10.1091/mbc.E18-01-0040)
For the 2019 human/mouse study, the complete normal abstract, MEF resorption Results and Figure 10 caption, early Discussion through proposed mechanisms, and patient/genetic/immunoblot/localization Methods were read, in addition to earlier targeted primary clinical/genetic passages. Arl2bp-null fibroblasts have shorter cilia and a delay at two hours after serum readdition; the resorption difference is not significant at six to 24 hours. Cell-cycle distribution was comparable. The authors retain uncertainty about direct fertilization, manchette mechanisms and a possible Sertoli-cell contribution. These results do not demonstrate intrinsic microtubule depolymerase activity. Complete remaining Methods, images and supplements were not audited. PMID:31425546(https://doi.org/10.1371/journal.pgen.1008315)
The complete 2013 normal abstract was checked against its identity and the earlier original-paper reading. The human nasal-cell localization result provides direct support for the existing basal-body annotation. The mixed-species binding constructs and separate HYDIN explanation in the other pedigree remain explicit. PMID:23849777(https://doi.org/10.1016/j.ajhg.2013.06.003)
The 50 original annotation objects and both alternative products are preserved. The prospective synthesis contains no new annotation. Its two activities are GTPase regulation and STAT3 binding with nuclear retention. The original 1999 paper establishes ARL2-GTP binding without GAP activity; the later 2021 study establishes cooperative ARL3 exchange and stabilization of the GTP-bound state. These experiments support regulation without assigning autonomous exchange-factor chemistry or a stable BART–ARL13B complex.
The current GO coactivator definition requires action via a DNA-bound transcription factor at a genomic locus. Reporter activity and nuclear retention do not alone demonstrate that mechanism. The original 2008 experiments support BART–STAT3 association and retention, with downstream transcriptional and phosphorylation effects distinguished from intrinsic DNA binding or kinase activity.
Independent primary-source consultation verified all four additional paper identities. The 2013 interaction experiment combines human Myc-ARL2BP, bovine VSV-ARL2 and human SK-N-SH cells; host species does not identify every construct. The second pedigree's primary ciliary dyskinesia/hearing findings have a separate HYDIN explanation and should not be assigned to ARL2BP. The mouse photoreceptor and sperm structural defects establish requirement for normal ciliary architecture without identifying an intrinsic microtubule polymerase activity. The molecular bridge between the reconstituted GTPase-regulatory activity and these structural phenotypes remains unresolved.
The GO-CAM index was searched for ARL2BP and Q9Y2Y0 and contained no matching entry at the time of this assessment. No process term is proposed from pathway completeness, and no new ciliary process assertion is inferred solely from a knockout phenotype. The uninspected source-specific interaction records and mitochondrial-matrix model details remain unresolved.
The earlier paragraphs preserve the chronology of the initial review. The reading below supersedes the earlier uncertainty about the two PMID:11847227 comparator experiments and strengthens the attachment of already-read STAT3 and localization results. All 50 original source annotations and both alternative products remain unchanged. Two binding records are refined, and one electronic localization mapping is marked unresolved; no NEW annotation or additional core function is introduced.
The original PMID:11847227 article text was recovered and read for Experimental Procedures/Plasmid Construction, Results/Binding of RP2 to Arl3, and the Figure 4A caption. The construct paragraph identifies human testis RNA as the source of the relevant cDNAs. Tagged recombinant ARLs were tested with radiolabeled, in-vitro-translated BART; BART is the positive control. The Results state “both Arl2 and Arl3 bound to BART”. Figure pixels and raw data were not audited. The normal cache remains abstract-only.
The original PMID:11847227 human BART positive-control pull-down experiments directly support association with both ARL2 and ARL3. They do not assign the RP2 tubulin-GAP activity to BART.
The current GO:0031267 definition is binding a small monomeric GTPase and places it under GTPase binding. It therefore describes these partner-specific results more usefully than protein binding. The UniProt reference annotation corroborates the intended citation; the direct experiment read resolves the prior source uncertainty.
The original author manuscript for PMID:18234692 was independently read for Methods pp.5–8 and Results pp.9–15, including the textual Figure 1–3 descriptions. The normal record remains abstract-only. Results/Figure 2C–D reports “vanadate treatment induced sustained tyrosine-phosphorylation of STAT3”; Figure 2A reports “IL-6/STAT3-mediated SOCS3 mRNA expression was markedly reduced”. These targeted reads do not establish the species of every tagged construct from its human host-cell line.
In the original PMID:18234692 Results, BART depletion reduces STAT3 tyrosine phosphorylation, and vanadate sustains phosphorylation in the depleted cells. This supports a regulatory contribution without establishing kinase activity of BART.
In the original PMID:18234692 Results, BART depletion reduces IL-6-induced SOCS3 mRNA in Hep3B cells and LIF-dependent reporter output in HeLa cells. These observations support context-dependent transcriptional regulation without demonstrating BART action at a genomic locus.
The original PMID:18234692 study includes endogenous BART–STAT3 co-immunoprecipitation in 293T cells and depletion/rescue experiments addressing STAT3 nuclear retention. This supports the existing STAT3-binding and retention function, separately from intrinsic DNA binding or kinase activity.
The phosphorylation and transcription rows now cite their relevant observations through these public notes, rather than an abstract sentence about nuclear retention. The original paper still does not establish locus-specific coactivator action.
The cached original PMID:16525022(https://pmc.ncbi.nlm.nih.gov/articles/PMC1446103/) localization Methods and targeted Results describe fixation conditions, antigen competition, preimmune controls and secondary-only controls. Weak but specific BART nuclear staining and the explicit conclusion of nucleoplasmic access now directly support the two nuclear-location rows. These data do not imply that the separate HPA images were inspected.
The actual SL-0251 mapping remains unresolved. The positive UniProt spindle location conflicts with its negative note and with spindle-fiber staining in the primary study. However, the current GO spindle definition includes spindle poles through part_of, and spindle-pole centrosome is beneath that pole class. The original study reports BART centrosomal throughout the cell cycle. Thus negative fiber staining does not establish that the broad GO component is incorrect, and the positive centrosomal pool does not by itself settle this particular mapping. Positive centrosome and midbody observations remain supported. The authors also allow possible midzone BART below antibody detection. No positive spindle-pole annotation, universal absence claim or NOT assertion is introduced. The current SL-0251 web body was unavailable during this check; the cited mapping provenance comes from the immutable seeded source.
The initial record contains 26 generic binding assertions: three already refined and 23 unresolved, not 26 unresolved binding assertions. Fourteen of the 23 involve ARL2/ARL3 and nine involve other partners. The two newly inspected human comparator records leave 21 source-specific interaction assertions unresolved. The three other UNDECIDED annotations concern mitochondrial-matrix Reactome models. Source availability flags describe the cached record, not whether a particular pair result was actually inspected.
The user-defined action criteria take precedence over a generic protein-binding removal default: REMOVE requires evidence against the assertion, while genuinely inaccessible relevant evidence remains UNDECIDED. Each remaining reason now states its specific source context and read limit. Independent binding biology is retained, but it is not silently substituted for an uninspected original assay. Missing target tokens do not establish absence of an experiment or its location in a supplementary table.
| Original source | Evidence actually inspected | Remaining scope |
|---|---|---|
| PMID:21988832 | Complete cached abstract and targeted screen Results. | Human liver interaction survey uses high-stringency yeast two-hybrid screening plus network integration. The ARL2BP–ARL2 record and any pair-specific validation were not inspected; this is not an endogenous liver-cell binding experiment merely because the survey concerns liver. |
| PMID:25416956 | Complete cached abstract and available Discussion; pair-level assay data uninspected. | The systematic human binary-interaction map is based on large-scale yeast two-hybrid work. Broad methods/context do not identify the configuration or validation of this exact ARL2BP–ARL2 record. |
| PMID:25502805 | Complete cached abstract and targeted Clone-seq, stability and interaction Results. | The study compares reference and disease-variant protein interactions and includes stability work in HEK293T. The exact ARL2BP–ARL2 reference/variant assay and its controls are unresolved. Neither a mutant-only result nor the absence of a reference interaction is inferred. |
| PMID:27107012 | Complete cached abstract and targeted BFG-Y2H principle, reference-set and centrosome-network Results. | Barcoded fusion-genetics yeast two-hybrid screens include reference sets and a centrosomal-protein matrix. The exact ARL2BP–ARL2 pair, which tested set it belongs to, and its score were not resolved. |
| PMID:27173435 | Complete cached abstract and targeted SF-TAP affinity-purification/mass-spectrometry network Results. | The ciliary network combines bait-prey and copurification relationships from HEK293T affinity purification. Exact ARL2BP–ARL3 and ARL2BP–CFAP20 records, association mode and scores remain uninspected. The network context is not automatically a direct binary binding assay. |
| PMID:29997244 | Complete cached abstract and targeted LuTHy Results and cellular assay Methods. | LuTHy combines live-cell BRET and luminescence coprecipitation after lysis. Which readout/configuration supports this ARL2BP–ARL2 record was not resolved; positivity in both modes is not assumed. |
| PMID:31515488 | Complete cached abstract and targeted yeast-two-hybrid variant-interaction Results. | The study compares missense-variant effects across human interaction pairs. The exact ARL2BP–ARL2 reference/variant profile and support were not inspected; no variant-specific functional failure is inferred. |
| PMID:32296183 | Complete cached abstract and targeted three-version yeast-two-hybrid screening Results. | HuRI combines repeated screens with different Y2H assay configurations. The exact ARL2BP records for ARL2, ARL3 and CFAP20, their assay versions and orientations, remain unresolved. |
| PMID:32814053 | Complete normal cached abstract; original full body/pair-level records unavailable in this read. | This neurodegeneration network combines targeted yeast-two-hybrid work and literature interactions. The six ARL2BP records cannot yet be assigned to an original assay or a literature-derived entry, and their exact target-specific outcomes remain unverified. |
| PMID:33961781 | Complete cached abstract and available Introduction/Discussion; source-specific pair results not inspected. | BioPlex compares affinity-purification/mass-spectrometry interactomes in 293T and HCT116. Exact ARL2BP–ARL2, ARL3 and CFAP20 bait/prey configurations, cell contexts and support were not resolved. The full_text_available flag does not mean these exact records were inspected. |
For PMID:26455799(https://pmc.ncbi.nlm.nih.gov/articles/PMC4635315/), the targeted original interaction, structural-comparison and Discussion passages principally describe new BARTL1–ARL3 work, with prior BART effector studies cited separately. The paper also reports variable BART ciliary localization in IMCD3 cells; it is therefore inaccurate to call every BART observation prior work. The specific ARL2BP–ARL3 IPI experiment is still unresolved, and no wrong-gene claim is made.
Both existing core functions are retained. The STAT3 association and retention experiments provide direct mechanistic observations in addition to perturbation effects. The ciliary genetics and ultrastructural phenotypes establish a requirement for normal architecture without identifying the exact step performed by ARL2BP. Accepted basal-body or centrosome localization does not by itself place the measured GTPase-regulatory reaction at those sites. No redundant binding core or new ciliary process is added.
The twelve remaining interaction records naming ARL2 (UniProtKB:P36404) or ARL3 (UniProtKB:P36405) now refine protein binding to small GTPase binding, GO:0031267. These are the records from PMID:21988832, PMID:25416956, PMID:25502805, PMID:26455799, PMID:27107012, PMID:27173435, PMID:29997244, PMID:31515488, and the two named partners in each of PMID:32296183 and PMID:33961781. The original curator already identified each partner; the refinement specifies that partner's molecular class, without adding an effector mechanism, direct binary interface, nucleotide preference or biological process. Independently established ARL2BP–ARL2/ARL3 binding supports retaining those named interactions. The shipped UniProt interaction list is corroboration, not proof that every accumulated observation occurred in every cited paper.
Each reason retains its source-specific read limit. In particular, variant profiles, bait/prey orientation, assay configuration, scores and cell context have not been newly resolved. The curated interaction can be retained without claiming those details were independently inspected. The nine non-GTPase interaction rows retain their source-local uncertainty. The earlier notes' counts describe the preceding review state; current totals are 15 ACCEPT, 4 KEEP_AS_NON_CORE, 17 MODIFY, 1 MARK_AS_OVER_ANNOTATED and 13 UNDECIDED across the same 50 original assertions. No NEW annotation is added; both products and both existing core functions are unchanged.
The spindle clarification concerns spindle pole centrosome, GO:0031616, a specialized structure that is part of spindle pole, GO:0000922, itself part of spindle. It does not place general centrosome, GO:0005813 universally within spindle. General centrosomes also exist in interphase. The original localization study reports BART at centrosomes throughout the cell cycle, but calls those BART all-stage images unpublished data; its controlled negative spindle-staining observation is retained. Together these findings leave the spatial scope of the SL-0251 electronic mapping unresolved. They neither establish a positive spindle-pole annotation nor exclude every structure within the broad GO component. Independently demonstrated centrosome localization is already retained in three existing annotations, so replacing the unresolved spindle mapping with another centrosome record would not resolve that mapping.
UNDECIDED applies when an annotation remains scientifically unclear; the user instruction additionally requires it when relevant evidence is inaccessible. Access to a paper does not eliminate ambiguity in the meaning of an electronic mapping. This follow-up retains the condition-specific negative observation, the positive centrosome/midbody observations, and the distinction between them. No universal NOT assertion, general-centrosome-to-spindle ancestry claim, new localization or new core edge is introduced.