DIS3L2 provides a cytosolic, exosome-independent route for RNA degradation. Its common molecular activity is divalent-metal-dependent, processive 3′–5′ RNA exonucleolysis. RNA recognition and cleavage of different substrates are coupled components of this integrated function; the core representation records recognition and exonucleolysis explicitly. Metal dependence is a supporting cofactor property. Structured-RNA degradation does not establish an independent ATP-dependent helicase function.
Lubas et al., PMID:23756462, Figure 2 and the recombinant-protein Methods, compare full-length human DIS3L2 with D391N. Purified protein degrades RNA from an accessible 3′ end, including partly structured substrates. Circular RNA and a blocked 3′ end resist cleavage in the tested assays. The β-globin/TNFα-ARE reporter in Figure 5 and selected endogenous-transcript half-life measurements in Figure 6 connect the enzyme to performed mRNA decay.
Figure 1C–D, the intracellular-localization Results and digitonin-fractionation Methods show tagged and endogenous human DIS3L2 distributed throughout the cytoplasm and recovered in its soluble fraction, separate from the nuclear/mitochondrial pellet. Figure 4 measures increased DCP1a- or XRN1-marked P bodies after DIS3L2 depletion. That is a downstream granule phenotype, distinct from localization of DIS3L2 itself to a P body. Cytosol is the supported specific location; this does not assert universal exclusion from granules under all conditions.
Figure 3A and the co-immunoprecipitation Results compare DIS3L2-FLAG with empty-host controls and RNA-preserving versus RNase-treated conditions. XRN1 is recovered by mass spectrometry and Western blotting in an RNA-dependent association. Its identity as an exonuclease does not turn RNA-dependent co-recovery into proof of a direct protein interface, an adaptor mechanism or an obligatory stable complex. Positive generic association is retained as non-core under the project curation instruction. The source's normal cache is abstract-only; the selected publisher Methods, Results, Discussion and figure captions were separately read.
Ustianenko et al., PMID:24141620, Figures 1–3, separates mouse embryonic-stem-cell oligo(U) pulldowns from human biochemical and cell experiments. Human protein binds U30 RNA, requires suitable divalent ions for cleavage, and loses measured cleavage after D391N substitution. Human-cell substrate trapping and purified FLAG preparations identify uridylated pre-let-7 as a substrate. Oligo(U) extensions promote degradation, but an A8 extension also promotes turnover; competition experiments show preference rather than an absolute uridylation requirement.
The Supplemental Methods, pages 6–8, identify the sequenced human construct as “isoform 1, 885 aminoacids, NP_689596.4”. The official UniProt record maps that exact RefSeq version to Q8IYB7-1. Bacterial His6-Smt3 protein underwent tag cleavage and gel filtration; the separate FLAG preparation came from HEK293T-Rex cells. This correspondence does not assign activity to the four shorter natural products. The main original and genuine supplement were inspected; the normal full-text flag is preserved.
GO:0010587 currently describes breakdown of mature approximately 21–23-nucleotide miRNA. The human study above sequences and degrades pre-let-7. Mature let-7 instead coprecipitates with AGO2 in the comparison, and DIS3L2 depletion in HeLa lowers mature let-7 without demonstrable precursor stabilization. These results do not invalidate direct precursor cleavage, but they cannot be described as direct mature-let-7 degradation.
Chang et al., PMID:23594738, Figures 1–4 and Methods, use an 870-residue mouse ESC-derived Dis3l2.2 construct, affinity-purified protein and bacterial recombinant protein with catalytic-mutant controls. Uridylated pre-let-7 is preferentially degraded. Mouse ESC depletion accumulates the precursor without raising the measured mature-let-7 level. The mouse structural study, PMID:25119025, measures oligo(U) binding and degradation and uses an engineered truncated D389N protein for crystallography. Its mouse constructs are distinct from natural human isoforms. These two studies ground precursor-RNA turnover, not a direct mature-miRNA assay.
There is separate mature-miRNA evidence. Lin et al., PMID:27974621, Figure 3B, the substrate-degradation Results and Methods, directly compare mature miR-1010 with its precursor using HA-Dis3L2 immunoprecipitated from Drosophila S2 cells. The separate Figure 3A wild-type/catalytic-dead comparison uses an artificial 77-nt substrate; Figure 3B compares mature and precursor substrates with wild-type immunoprecipitates. The control supports the nuclease assignment without being a matched mature-miRNA-panel control. Mature-miRNA degradation therefore has genuine fly-donor grounding for the human phylogenetic annotation, even though endogenous human mature-miRNA substrate scope remains less clear. All four miRNA-catabolism rows retain the same non-core gene-level verdict on this combined evidence. The human and mouse experiments keep their precursor-specific scope, without a redundant broader RNA-catabolism replacement. No precursor-specific GO identifier is invented.
Malecki et al., PMID:23503588, the localization, transcript-tail and biochemical Results with their Methods, concern Schizosaccharomyces pombe O14040/SPAC2C4.07c. Endogenous GFP-tagged protein has a cytosolic pool and foci adjacent to Dcp2-positive P bodies; glucose deprivation increases colocalization. Genetic interaction with other decay pathways, uridylated trimmed adh1 transcripts in the double-mutant context and purified-protein substrate preference support uridylation-coupled mRNA decay. They are not human localization measurements. The normal cache is abstract-only, with the selected original separately accessed.
Zhao et al., PMID:34904664, Figure 4 and the mRNA-tail Results, show that DIS3L2 depletion stabilizes maternal transcripts and enriches oligo(U)-tailed Gm4745 species during mouse early embryonic decay. Separate tagged-protein HeLa fractionation and interaction experiments should not be conflated with the mouse embryo assays. DIS3L2 performs RNA destruction downstream of uridyltransferases; it does not add the U tail.
The mouse let-7 paper establishes precursor decay in ESCs and a connection to the pluripotency pathway. Its inspected assays do not demonstrate the self-renewal or niche-maintenance step described by stem-cell population maintenance. That specific transferred process is graded as over-annotated relative to the demonstrated RNA-decay role, without denying developmental consequences of DIS3L2 loss.
Astuti et al., PMID:22306653 explicitly reports exonuclease activity, mitotic abnormalities, altered checkpoint-protein expression and growth effects in its abstract. The original full text was not recovered through the attempted publisher and public-author routes. Broad mitotic and proliferation effects are retained as non-core with curator deference. The specific sister-chromatid-separation assay remains unresolved; no source misattribution is inferred from the access gap.
The PTHR23355 tree places human Q8IYB7 at PTN002511618. The actual path contains all four imported ancestral nodes: PTN000599170 for exonuclease activity, PTN000599171 for mRNA decay, PTN000599278 for the imported P-body assertion and PTN001156591 for miRNA decay. The current PAINT IBD file, inspected on 2026-10-17, now assigns cytosol rather than P-body at PTN000599278. Imported GOA source fields remain unchanged. Target inclusion among descendant evidence is expected and is not treated as circularity.
Current experimental donor joins are directly inspectable for mouse Q8CI75, fission-yeast O14040 and fly Q8IRJ7. The fly IDA for GO:0010587 cites PMID:27974621. Donor number is not used as a measure of phylogenetic support.
IPR001900 matches the human RNB region at residues 371–721; IPR028591 matches the DIS3L2 family region at residues 17–861. UniRule UR000187751 requires the HAMAP family, eukaryotic taxonomy and nonfragment status, with additional taxonomic rules. ARBA00027832 has a matching IPR001900/IPR022966/Chordata condition. The full magnesium, poly(U), miRNA and cytoplasm rules were preserved, but no exact human FunFam match is claimed where it was not reconstructed. Direct experiments independently support the accepted biochemical properties.
Official alternative-sequence records reconstruct products of 885, 619, 603, 249 and 356 residues. The four shorter products have distinct C-terminal changes. The human 885-residue assay correspondence is established above; no separate functional assay for the shorter natural products was established in the inspected source set. Sequence differences alone do not demonstrate equal function or measured inactivity. No new annotation or functional-isoform class is proposed.