ELP2 is a WD40-containing component of Elongator. Human ELP4-FLAG purification identifies endogenous ELP2 in the assembled complex, and human tagged ELP2 N-terminal deletion reduces association with ELP1/ELP3. These experiments support an assembly contribution; they do not establish an independent ELP2 enzyme (PMID:22854966(https://pmc.ncbi.nlm.nih.gov/articles/PMC3463322/), Experimental Procedures, Affinity Purification; Results, Identification of the Components of the Human Elongator Complex, Fig. 1A; PMID:25960406(https://www.researchgate.net/publication/276148381_The_Elp2_Subunit_Is_Essential_for_Elongator_Complex_Assembly_and_Functional_Regulation), Results, human N17 deletion/co-immunoprecipitation, Fig. 3C).
The conserved tRNA wobble-uridine modification role is supported by the ELP2-specific InterPro family, Elongator membership and the reviewed literature. The human record describes cm5U formation by Elongator. The chemical activity of the machinery must remain distinct from the activity of ELP2 alone (PMID:27847465(https://pmc.ncbi.nlm.nih.gov/articles/PMC5088202/), Elongator organization and substrate-specificity sections; UniProtKB:Q6IA86, Function). Yeast structural, microtubule-binding and H3K14 results are not rephrased as direct human ELP2 assays.
The 20 imported annotations remain 20 separate source objects and match all 20 normal GOA contexts. Actions are 7 ACCEPT, 6 MODIFY and 7 KEEP_AS_NON_CORE. The three cytoplasm-to-cytosol and three nucleus-to-nucleoplasm refinements consistently use independent target-specific HPA evidence; they do not claim that the original antibody experiments or SL mapping resolved those compartments. Cytosol is the main location and nucleoplasm is additional. No exclusive localization or PML-body subtype is inferred.
Historical RNA polymerase II association and transcription experiments require separate interpretation. PMID:11714725(https://www.researchgate.net/publication/11640079_Purification_and_Characterization_of_the_Human_Elongator_Complex) assayed the human complex with calf-thymus polymerase (RNA polymerase II interaction Methods and Fig. 2), associated it with early elongation complexes, and localized ELP1/ELP3 (Immunostaining Methods and Fig. 3). It did not perform the transcription-rescue experiment previously attributed to it in the review. PMID:11818576(https://pmc.ncbi.nlm.nih.gov/articles/PMC122174/) directly included anti-Elp2 immunofluorescence (Methods, Immunofluorescence; Results, Cellular Localization; Fig. 3). Its Functional Analysis and Fig. 5 report acetyl-CoA-dependent rescue of chromatin transcription in Elongator-depleted nuclear extract, despite failed stimulation in a fully defined reconstituted system. The Discussion leaves a possible initiation contribution open. These are positive complex-level observations with limited mechanistic resolution, not isolated ELP2 catalysis.
The predominant modern function is tRNA U34 modification. PMID:27847465(https://pmc.ncbi.nlm.nih.gov/articles/PMC5088202/), substrate-specificity discussion and Conclusions, explains how tRNA rescue can bypass some whole-cell transcription-related phenotypes in yeast, while leaving distinct transcription functions possible. Secondary translation defects cannot simply be assumed to explain purified polymerase binding or acute nuclear-extract add-back. Consequently, the four imported transcription-era annotations remain non-core on their positive biochemical and complex-association evidence, with the uncertain physiological and transcription-step specificity stated explicitly. The imported contributes_to qualifier on polymerase binding remains intact.
The protein-kinase-binding transfer has a literal mouse Q91WG4 experimental donor (PMID:10954736(https://pmc.ncbi.nlm.nih.gov/articles/PMC27739/), Results, StIP1 Associates with JAKs, Fig. 4; trimolecular-complex tests, Fig. 5). Murine full-length StIP1, its WD7-deleted form and GST/HA fragments must remain distinct from the seven current human products. Positive JAK association does not establish a stable ternary STAT3–ELP2–JAK scaffold, which the inspected study did not recover.
Seven official products are preserved with their original IDs, names and sequence notes. Their reconstructed lengths are 826, 756, 800, 705, 821, 891 and 756 aa. Historical and engineered constructs are not new splice products, and no equal activity across the seven is claimed. Sequence-only resources and clinical abstracts were assessed privately without converting them into additional molecular or developmental assertions.
The source guide distinguishes immutable normal caches from separately inspected originals. Four selected original papers have abstract-only/partial normal caches despite bounded external original access; optional false full_text_unavailable keys are omitted. PMID:29332244 remains abstract-only and is flagged accordingly. The provider attempt failed genuinely; no provider-labelled substitute was authored. No new publication quotations, source defects or annotation rows were added.
Reproducible database and sequence checks are in ELP2-bioinformatics/RESULTS.md. The visible script verifies input integrity and regenerates results from database snapshots. Original article bodies, figures and provider logs are excluded; their public routes, hashes, inspected sections and limits are listed in the source guide and original-locators.json.
Direct original URLs and named experimental sections now accompany the biological claims above and in each publication assessment. These assessments distinguish separately accessible original text from the immutable abstract-only normal caches. No new publication quotations or cache edits were made. PMID:29332244 remains an abstract-only background source; the full-text modern reinterpretation used here is PMID:27847465. GO method references are graded as contextual rather than primary functional evidence.
The independent HPA record supports both compartment refinements consistently. Its HPA002996 and CAB020835 antibody records include nucleoplasmic observations in named human cell lines; their main cytosolic classification does not exclude a smaller nuclear pool. The review cites the existing reproducible report for all six refinements. Human microtubule binding and the direct-versus-secondary origin of nuclear transcription effects are now explicit questions. The original kinase-binding donor remains positively supported: absence of a stable ternary signaling complex does not negate the observed binary JAK interactions. All 20 imported source objects, seven natural products and the structural contribution to the tRNA-modification core are preserved.