DIS3L2

Summary: Conserved 3′–5′ RNA exonuclease activity is directly confirmed in human DIS3L2.

Reason: The human leaf PTN002511618 descends from the experimentally grounded IBD node PTN000599170 in PTHR23355. Human recombinant-protein assays in PMID:23756462 and PMID:24141620 independently establish cleavage from free RNA 3′ ends with catalytic-mutant controls. Target self-inclusion among descendant evidence is expected; it is not circular support. No loss on the target lineage is established.

Summary: Purified human DIS3L2 is a processive 3′–5′ exoribonuclease.

Reason: PMID:23756462 tests full-length human recombinant wild-type and D391N proteins, 5′-labelled RNA substrates, circular RNA and blocked 3′ ends. Cleavage requires the catalytic residue and accessible RNA 3′ termini; structured substrates can also be degraded. This is the measured core activity, without inferring an independent helicase or DNA-nuclease activity.

Summary: Human isoform-1 DIS3L2 catalyses RNA exonucleolysis.

Reason: PMID:24141620 distinguishes initial mouse oligo(U) discovery from subsequent human isoform-1 biochemical experiments. Recombinant human protein degrades U30 RNA to short products, whereas D391N lacks cleavage. The supplement identifies the 885-residue NP_689596.4 construct; these measurements are not isoform-resolved evidence for the four shorter natural products.

Summary: The electronic nuclease assignment agrees with direct human biochemistry.

Reason: IPR028591 matches human DIS3L2 residues 17–861 and maps to GO:0000175. UniRule UR000187751 has the DIS3L2 family condition and the same conclusion. ARBA00027832 includes an IPR001900 plus IPR022966 plus Chordata condition that matches the observed human entries and taxonomy. Mouse Q8CI75 is experimentally characterized, and human catalytic assays provide independent confirmation.

Summary: DIS3L2 affects mitotic progression through its RNA-metabolism role.

Reason: The PMID:22306653 abstract reports mitotic abnormalities and altered checkpoint-protein expression following DIS3L2 inactivation. Retain the curated experimental connection as a contextual cellular consequence of RNA turnover; this does not establish DIS3L2 as a chromosome-segregation motor or a separate cell-cycle catalytic unit. The original full assay details were not accessible, so no specific construct or mitotic-stage mechanism is inferred.

Summary: Magnesium supports the human DIS3L2 catalytic reaction.

Reason: PMID:24141620 Figure 2E tests divalent-ion dependence and EDTA inhibition of human DIS3L2 RNA degradation. This supports the curated magnesium-binding cofactor annotation in the integrated nuclease mechanism. It is an ion-dependent biochemical assay, not a direct human metal-site structure or an independently functioning metal-binding unit. Retain this cofactor-binding property as non-core rather than represent it as a separate evolved metal-binding activity.

Summary: The magnesium-cofactor inference is consistent with conserved nuclease chemistry.

Reason: The mouse Q8CI75 donor has IDA evidence from PMID:25119025, which assays Mg-dependent RNA degradation; current ARBA00028486 concludes magnesium binding. Direct human ion-dependence experiments in PMID:24141620 corroborate the biological claim. The exact CATH FunFam condition branch is not reconstructed here, and no residue-level metal geometry is transferred. Retain this cofactor-binding property as non-core rather than represent it as a separate evolved metal-binding activity.

Summary: Refine the inherited P-body location to the directly supported cytosol.

Reason: The human leaf is inside PTN000599278, but the current IBD file places cytosol, not P-body, at that node. Fission-yeast Dis3l2 shows P-body-adjacent or glucose-starvation-colocalizing foci in PMID:23503588. In contrast, PMID:23756462 directly finds tagged and endogenous human DIS3L2 uniformly distributed in the cytoplasm and enriched in the digitonin-soluble cytoplasmic fraction. Preserve the imported P-body source row while replacing its location with cytosol; this is not a claim that human DIS3L2 can never enter a granule.

Proposed replacements: cytosol

Summary: The cited human study directly supports cytosol rather than a P-body localization.

Reason: PMID:23756462 Figure 1C–D and the fractionation Methods demonstrate homogeneous endogenous/tagged human cytoplasmic distribution and a digitonin-soluble pool. The P-body experiments instead measure DCP1a/XRN1-marked granule increases after DIS3L2 depletion. That depletion phenotype is not a localization assay for DIS3L2 in P bodies. Refine to the measured cytosol without asserting universal granule exclusion.

Proposed replacements: cytosol

Summary: Refine the broad family-level P-body prediction to human cytosol.

Reason: UniRule UR000187751 explicitly includes P-body among its family-wide location conclusions; the related fission-yeast phenotype is experimentally real. Human endogenous localization and soluble fractionation in PMID:23756462 provide a better supported target location. The current evidence does not justify importing the yeast granule context as the preferred human location.

Proposed replacements: cytosol

Summary: DIS3L2 directly contributes to degradation of nuclear-transcribed mRNAs.

Reason: The UniRule assignment is supported independently by human β-globin/ARE reporter decay and selected endogenous-mRNA half-life measurements in PMID:23756462, together with direct exonuclease chemistry. Nuclear-transcribed describes RNA origin, not the cellular location of degradation.

Summary: Human DIS3L2 promotes decay of selected mRNAs.

Reason: PMID:23756462 uses a tetracycline pulse–chase β-globin/TNFα-ARE reporter and metabolic labelling of endogenous transcripts. DIS3L2 depletion prolongs their half-lives. Combined with the purified-enzyme assays, this establishes performed RNA degradation rather than merely a secondary change in transcript abundance.

Summary: Use the directly demonstrated poly(U)-RNA-binding specificity.

Reason: The IPR001900 RNB-domain mapping correctly identifies RNA interaction but is broad. Human recombinant DIS3L2 binds U30 RNA with nanomolar affinity in PMID:24141620; competition experiments identify an oligo(U) preference. GO:0008266 captures this demonstrated specificity without claiming that other RNA substrates cannot bind or be degraded.

Proposed replacements: poly(U) RNA binding

Summary: Refine RNA nuclease activity to established 3′–5′ RNA exonuclease activity.

Reason: The PMID:22306653 abstract explicitly reports exonuclease activity, and independent human biochemical studies establish direction, accessible-3′-end requirement and D391N dependence. The original 2012 full methods were inaccessible; the precise refinement uses the combined evidence and does not invent its assay details.

Proposed replacements: 3'-5'-RNA exonuclease activity

Summary: Refine the RNB-domain nuclease mapping to its measured reaction.

Reason: IPR001900 matches residues 371–721 and carries RNA-nuclease and RNA-binding mappings. Human DIS3L2 experiments identify processive 3′–5′ exonucleolysis, while tested circular and 3′-blocked RNA substrates lack the alternative cleavage patterns. GO:0000175 is the supported specific activity.

Proposed replacements: 3'-5'-RNA exonuclease activity

Summary: Retain the positive RNA-dependent association with XRN1.

Reason: PMID:23756462 identifies human XRN1/Q8IZH2 with DIS3L2-FLAG by co-IP/MS and Western blotting, with empty-host and RNase-sensitive comparisons. XRN1 is an exonuclease, but RNA-dependent co-recovery does not establish a purified protein interface or justify an enzyme-binding, adaptor or obligate-complex mechanism. Retain this contextual association under the project generic-binding policy rather than discarding positive evidence.

Summary: Refine cytoplasm to the demonstrated soluble cytosolic pool.

Reason: The PMID:22306653 source is retained; its full original localization assay was inaccessible. The specificity refinement relies independently on PMID:23756462. Endogenous human DIS3L2 is recovered in the digitonin-soluble cytoplasmic fraction, separated from the nuclear/mitochondrial pellet, with homogeneous cytoplasmic imaging. Cytosol is already represented elsewhere in the seed; this replacement adds precision, not another biological function.

Proposed replacements: cytosol

Summary: Refine cytoplasm to the demonstrated soluble cytosolic pool.

Reason: PMID:23756462 directly supplies the endogenous/tagged localization and fractionation evidence. Endogenous human DIS3L2 is recovered in the digitonin-soluble cytoplasmic fraction, separated from the nuclear/mitochondrial pellet, with homogeneous cytoplasmic imaging. Cytosol is already represented elsewhere in the seed; this replacement adds precision, not another biological function.

Proposed replacements: cytosol

Summary: Refine cytoplasm to the demonstrated soluble cytosolic pool.

Reason: PMID:24141620 directly localizes human DIS3L2 in HeLa/HEK293 cells; the soluble-subcompartment refinement is independently grounded in PMID:23756462. Endogenous human DIS3L2 is recovered in the digitonin-soluble cytoplasmic fraction, separated from the nuclear/mitochondrial pellet, with homogeneous cytoplasmic imaging. Cytosol is already represented elsewhere in the seed; this replacement adds precision, not another biological function.

Proposed replacements: cytosol

Summary: Refine cytoplasm to the demonstrated soluble cytosolic pool.

Reason: The pooled electronic cytoplasmic assignment is consistent with mouse donor and UniRule records; independent human experiments resolve the compartment more precisely. Endogenous human DIS3L2 is recovered in the digitonin-soluble cytoplasmic fraction, separated from the nuclear/mitochondrial pellet, with homogeneous cytoplasmic imaging. Cytosol is already represented elsewhere in the seed; this replacement adds precision, not another biological function.

Proposed replacements: cytosol

Summary: Cytosol is directly supported for human DIS3L2.

Reason: The imported mouse-based IEA is independently supported by endogenous human microscopy and digitonin-soluble fractionation in PMID:23756462. The annotation does not imply that all natural products were separately localized or that localization is exclusive under every condition.

Summary: Refine the conserved mRNA-decay process to nuclear-transcribed mRNA decay.

Reason: PTN000599171 is a real ancestral node on the human lineage with experimental descendant grounding, including human DIS3L2. Human reporter and endogenous-transcript turnover experiments identify nuclear-transcribed mRNAs degraded in the cytoplasm. The more precise existing GO:0000956 describes that substrate context; self-inclusion and the number of donors are not defects.

Proposed replacements: nuclear-transcribed mRNA catabolic process

Summary: Human DIS3L2 directly binds oligo(U) RNA.

Reason: PMID:24141620 Figure 2D uses recombinant human protein in an electrophoretic mobility-shift assay with U30 RNA; competition and substrate assays establish oligo(U) preference. This is a substrate-recognition component of the exonuclease function, not a separate regulatory role.

Summary: Conserved poly(U)-RNA recognition has direct human support.

Reason: Mouse Q8CI75 has IDA evidence from PMID:25119025, including quantitative RNA binding and an oligo(U)-bound structure of an engineered mouse protein. ARBA00092390 concludes poly(U) binding. Human U30 binding in PMID:24141620 independently validates the target claim; no exact mouse construct or unverified FunFam match is assigned to a natural human product.

Summary: Retain the experimentally supported antiproliferative context.

Reason: PMID:22306653 reports that DIS3L2 overexpression suppresses growth of human cancer cell lines and knockdown enhances growth. This supports the curated negative-regulation annotation in those experimental contexts. It does not establish a universal antiproliferative effect across tissues or a distinct catalytic activity; the full original assay details were inaccessible.

Summary: Retain the phylogenetic mature-miRNA-decay claim with its donor scope.

Reason: The human lineage descends from PTN001156591. Current FlyBase IDA grounding from PMID:27974621 directly compares mature miR-1010 and precursor degradation by wild-type Drosophila HA-Dis3L2 immunoprecipitates in Figure 3B. The separate Figure 3A catalytic-dead comparison uses an artificial 77-nt substrate, supporting nuclease specificity without being a matched control for the mature-miRNA panel. This supports inherited catalytic capacity. Human PMID:24141620 and mouse PMID:23594738/25119025 instead test precursor let-7, so they are not direct mature-miRNA degradation demonstrations. Retain the valid phylogenetic claim as non-core pending clearer endogenous human mature-miRNA substrate evidence; do not call target self-evidence circular.

Summary: Retain the gene-level mature-miRNA-decay claim while distinguishing the cited precursor assay.

Reason: PMID:24141620 directly tests uridylated pre-let-7, sequences the trapped precursor, and finds mature let-7 associated with AGO2 in the comparison. Its HeLa depletion result lowers mature let-7 rather than demonstrating its degradation. Those source limits remain important, but the combined gene-level evidence includes direct mature-miR-1010 degradation by the fly ortholog in PMID:27974621 Figure 3B and a valid human descendant of the PTN001156591 ancestral assertion. Retain the same non-core verdict across all four miRNA-catabolism rows. No broader RNA-catabolism replacement is added, because it is already entailed by the retained claim.

Summary: Retain the electronic mature-miRNA claim with source-specific uncertainty.

Reason: ARBA00092385 concludes miRNA catabolism through several family-condition branches; the exact human FunFam match is not independently reconstructed. UniRule and mouse references emphasize pre-let-7, which is narrower experimental context than the current mature-miRNA definition. Independent mature-miRNA degradation in the fly donor PMID:27974621 and the valid PTN001156591 placement make the gene-level claim defensible, while direct human mature-miRNA physiology remains insufficiently resolved for core designation.

Summary: Retain the shared miRNA-decay claim without relabelling mouse precursor assays as mature-miRNA experiments.

Reason: The literal Q8CI75 donor joins GO:0010587 to PMID:23594738 and PMID:25119025. These experiments test precursor let-7 and oligo(U) substrates, and the 2013 study reports unchanged mature let-7 after mouse ESC depletion. The combined target claim is nevertheless supported by the independently verified fly mature-miR-1010 experiment in PMID:27974621 and the valid PTN001156591 inheritance. Retain the gene-level claim as non-core, consistently with the other rows, while preserving the mouse-source caveat. A broader RNA-catabolism replacement would be redundant with the retained process.

Summary: The stem-cell maintenance claim exceeds the directly demonstrated precursor-decay role.

Reason: Mouse Q8CI75 is annotated from PMID:23594738. Its operative experiments establish uridylated pre-let-7 degradation in ESCs, with unchanged mature let-7 after depletion; the paper links the pathway to pluripotency but does not demonstrate a stem-cell population-maintenance step in these assays. GO:0019827 specifically concerns maintaining a population through mechanisms such as self-renewal or niche maintenance. This is a plausible indirect pathway connection, but its transfer overstates the experimentally established RNA-decay role; no claim of universal absence of developmental function is made.

Summary: The stem-cell maintenance claim exceeds the directly demonstrated precursor-decay role.

Reason: Mouse Q8CI75 is annotated from PMID:23594738. Its operative experiments establish uridylated pre-let-7 degradation in ESCs, with unchanged mature let-7 after depletion; the paper links the pathway to pluripotency but does not demonstrate a stem-cell population-maintenance step in these assays. GO:0019827 specifically concerns maintaining a population through mechanisms such as self-renewal or niche maintenance. This is a plausible indirect pathway connection, but its transfer overstates the experimentally established RNA-decay role; no claim of universal absence of developmental function is made.

Summary: The specific sister-chromatid separation assignment remains unresolved.

Reason: The accessible PMID:22306653 abstract reports mitotic abnormalities after DIS3L2 inactivation but does not specify the chromosome-detachment assay needed to assess GO:0051306. The original full text could not be recovered through the attempted publisher and public-author routes. Preserve the curator’s experimental assertion pending direct inspection rather than infer a misattribution or a chromosome-separation mechanism from the abstract.

Summary: The conserved uridylation-coupled mRNA decay route is biologically supported.

Reason: The mouse Q8CI75 donor has direct perturbation evidence in PMID:34904664: DIS3L2 depletion stabilizes maternal transcripts and enriches oligo(U)-tailed Gm4745 species during early embryonic decay. UniRule assigns the same pathway, and fission-yeast and human biochemical studies establish oligo(U)-preferential exonucleolysis. DIS3L2 performs RNA destruction; upstream TUT enzymes add the U tail. This does not imply that every human substrate must be uridylated.

Summary: Retain the fission-yeast-supported uridylation-dependent mRNA-decay transfer.

Reason: O14040 is S. pombe Dis3l2/SPAC2C4.07c, not a mouse protein. PMID:23503588 combines genetic mRNA-turnover evidence, enrichment of uridylated trimmed adh1 transcripts in the double-mutant context and purified-enzyme substrate preference. The conserved human enzyme directly recognizes and cleaves oligo(U)-tailed RNA. These complementary data support the transferred degradation pathway while preserving organism and genetic-context limits; DIS3L2 is not the uridyltransferase.