DNASE1L3 hydrolyzes DNA rather than merely being required for DNA disposal. Human recombinant-enzyme work supports the DNase I chemistry, and human expression-supernatant and patient-plasma assays support digestion of nucleosomal and microparticle-associated DNA. The latter assays are not purified-enzyme kinetics. Engineered C-terminal truncations and tags help define substrate access but are not the natural internally deleted product. PMID:9070308(https://pubmed.ncbi.nlm.nih.gov/9070308/), PMID:11141064(https://pubmed.ncbi.nlm.nih.gov/11141064/), PMID:9714828(https://pubmed.ncbi.nlm.nih.gov/9714828/), PMID:27293190, Results/Figure 5 and recombinant-DNase Methods.
The intracellular evidence is retained with its experimental context. In the recovered original, human DNAS1L3 was expressed in mouse MEFs and compared with ICAD-deficient cells. The endogenous localization experiments instead used human HT-29 cells. Figure 3 explicitly states, “Cells were also stained with ER tracker and DAPI.” Fractionation also used compartment controls. These observations support ER and inducible nuclear pools without resolving an ER membrane/lumen or nuclear subcompartment. PMID:23229555, Methods and Figures 2–3. The separately recovered author manuscript and exact public section locators are listed in the source guide; the immutable normal cache is abstract-only.
Conditional fragmentation does not establish a universal apoptotic requirement. Later human cell-line experiments distinguished necrotic from CAD-dependent apoptotic fragmentation, and mouse knockout tissues retained fragmentation in the tested apoptotic conditions while losing the necrotic response. Those contexts qualify the earlier positive experiments rather than erase them. The inherited developmental annotation has positive mouse C2C12 donor grounding and is retained as non-core. PMID:24312463, Results/Figures 1–4 and Methods, PMID:12050166(https://pubmed.ncbi.nlm.nih.gov/12050166/).
The inspected NET study establishes digestion after NET release. Its mouse experiments and human clinical observations remain distinct. The accessible main article supports an inflammatory regulatory context, while neutrophil activation and directed target-cell killing remain unresolved under their precise GO definitions; supplementary methods were unavailable. PMID:29191910, official digital article pp. 1202–1206.
The complete PTHR11371 tree and alignment place human Q13609 below the imported PTN000889801 and PTN000146474 nodes. The aligned target exactly matches the official 305-residue sequence. Literal IBD records and current experimental donor records were traced; neither a short donor list nor the target's own experimental contribution was treated as a defect. The rat DNA-binding/nuclear evidence is positive at abstract and curator-record level, with its full methods unavailable. PMID:9665719(https://pubmed.ncbi.nlm.nih.gov/9665719/). Current ARBA00027850 branch 19 matches the target's actual FunFam and primate lineage; domain/rule matches are provenance rather than independent assays.
HuRI pairs resolve to CYSRT1 and KRTAP5-9. Exact human sequences match the archived ROGIDs, with array, pooling and validated-two-hybrid records. These show tagged binary binding capacity in yeast, not a demonstrated physiological nuclease complex. Generic binding is retained as non-core under the scoped project instruction. No assembled intermediate-filament binding is inferred from the keratin-associated partner's name. PMID:32296183(https://pmc.ncbi.nlm.nih.gov/articles/PMC7169983/).
Official products are 305 and 275 residues. The shorter product deletes displayed residues 78–107. The annotated signal peptide is 1–20 and mature chain 21–305; no newly inspected processing assay is claimed. The natural short product's activity and trafficking remain unresolved. Conserved calcium-binding residues and human calcium-dependent catalysis support the cofactor annotation without establishing a separate calcium-sensing function or a measured affinity. PMID:9205125(https://pubmed.ncbi.nlm.nih.gov/9205125/), UniProt Q13609.
All 31 fetched annotation source objects are preserved. One extracellular catalytic core integrates the principal activity; conditional intracellular and immune contexts remain separately explained. No NEW annotation is proposed. The genuine Falcon request failed authentication (HTTP 403); the actual perplexity-lite fallback then failed quota (HTTP 401). No provider report was fabricated. Normal publication caches retain their actual access flags; independently recovered originals are documented separately. Saved-only source, sequence, propagation and quote checks accompany the delivery; final ordinary validation and rendering are performed by the canonical publisher.
This entry supersedes the earlier uncertainty-only NET verdicts and the extracellular-only core summary. The 2017 original, Results and Figures 1–4 demonstrates digestion of already released NETs and protection from obstructive NET clots. G-CSF induces neutrophilia; circulating DNase clears the extracellular substrate. That step does not establish neutrophil activation or directed target-cell killing, so the four transferred annotations are now marked over-annotated. The inaccessible supplementary Methods remain an access limit, not a reason to ignore the demonstrated mechanism. The broad inflammatory-regulation rows remain non-core: less vascular injury during continued neutrophilia does not by itself resolve a narrower negative-regulation step.
Current GO searches found NET formation and symbiont-mediated NET disruption, but no matching host NET-disassembly term. A proposed ontology term records that gap without adding a NEW annotation. Neither formation nor suppression of formation is substituted for degradation. The extracellular core now explicitly includes released NET DNA alongside microparticle-associated chromatin.
A second catalytic context records cleavage of nuclear chromatin in dying cells, using existing GO:0004530 and GO:0006308. PMID:24312463, Figures 1–4 and Cells and plasmid constructs supplies positive human-DNase transgene results in human cell lines, separately corroborated by mouse knockout and ischemia experiments. PMID:23229555, Figure 3 and fractionation Methods establishes the induced endogenous nuclear pool in human HT-29 cells. The four nucleus annotations now support this core location. Apoptosis-specific assertions remain non-core because CAD cooperation and DNASE1L3 dependence vary with the tested context. The route by which nuclease reaches injured-cell chromatin remains unresolved; a nuclear activity site does not imply constitutive nuclear residence or initiation of necrosis.
The two newly fetched sources add direct biochemical support. PMID:14646506(https://doi.org/10.1023/A:1009692807692) reports human Trx-DNase gamma chromatin cleavage with Ca/Mg or Mn dependence. PMID:10807908(https://doi.org/10.1074/jbc.M001087200) reports recombinant human nuclease cleavage and in-vitro PARP modification with loss of activity; its transfected cellular experiments use mouse skin fibroblasts. Both normal caches are abstract-only, and bounded publisher attempts did not recover full Methods. No liposome assay is attributed to either abstract. Microparticle substrate evidence remains grounded in the independently read PMID:27293190 original.
GO:0004530 describes DNA cleavage chemistry, not identity with the DNASE1 gene or susceptibility to G-actin. Its existing phylogenetic assertion and biochemical grounding are retained; UniProt's partial EC number does not negate the reaction class. The inspected studies do not establish every end-product distribution independently. The natural 275-residue product remains distinct from engineered terminal deletions and tagged constructs; historical assays are not automatically assigned to every splice product. Product descriptions now state biology, with these interpretation limits retained here. All 31 imported source dictionaries, both product identities and existing caches are unchanged.
Porat et al. 2026 (Mol Cell 86:4291–4309) report that mitochondrial DNA is tethered to the surface of select mammalian cells through cell surface proteins and heparan sulfate proteoglycans, and that DNASE1L3 clears it PMID:42838054. The published version is abstract-only here; the open preprint has the full text.
What the preprint shows:
- Surface DNA released by trypsin is mitochondrial and oxidized PMID:41280093. This is in cell lines; primary B-cell surface DNA was imaged, not sequenced.
- Among primary immune cells it is on B cells and monocytes PMID:41280093.
- Recombinant enzyme removes it, and DNASE1 does not PMID:41280093.
- R206C (rs35677470) homozygotes have more PMID:41280093. R206C is secretion-defective PMID:36708011.
Decisions:
- No NEW annotation. The right process term, GO:0006308 DNA catabolic process, is already in GOA (IEA), so this paper is added as experimental support on that accepted row and in a separate third core function. DNASE1L3 cuts the DNA itself, so this is participation, not a substrate claim.
- GO:0032043 mitochondrial DNA catabolic process is not used. Its text definition reads as substrate-only, but its logical definition is GO:0006308 ⊓ occurs_in GO:0005739 mitochondrion. "Mitochondrial" names where the breakdown happens, so applying it here would wrongly place DNASE1L3's activity inside mitochondria. (An earlier commit on this branch used it; corrected after curator review.)
- Location stays GO:0005576 extracellular region. DNASE1L3 is secreted and acts on surface DNA from outside; nothing shows the protein residing at the cell surface, so GO:0009986 cell surface is not added. GO:0005615 extracellular space is now obsolete in GO.
- No immune-process term such as tolerance: the anti-DNA response in deficient mice is carried out by B cells, pDCs and TLRs PMID:32454024, not by DNASE1L3.
- Proposed new term "extracellular DNA catabolic process". GO cannot say where DNA is degraded, and that location is the whole point of DNASE1L3: plasma DNA PMID:30593563, microparticles, NETs and now the cell surface.
Rows: GO:0005783 IDA (PMID:23229555) and IEA (GO_REF:0000044, SL-0095), both located_in; no ER sub-compartment rows.
Two blocking items, both taken. The first is the one worth recording, because the file already contained the correction and I ignored it.
The B-cell result was given a substrate identity the papers do not give it. The top-level description and the second core function both said DNASE1L3 removes oxidized mitochondrial DNA from the surface of B cells. The published abstract says something weaker and more careful PMID:42838054 and, separately, PMID:42838054 — DNA, not mitochondrial DNA, and no oxidation claim attached to the B-cell experiment. The mitochondrial identity and 8-oxo-dG enrichment come from cell lines PMID:41280093. My own reference_review for the preprint had already recorded this — "Mitochondrial identity of surface DNA was shown in cell lines, not sequenced from primary B cells" — so the two most reader-facing fields in the file contradicted a caveat sitting a few hundred lines above them. Both now separate the two claims and say which rests on which.
The second core function was machine-identical to the first. Same GO:0004530, same GO:0006308, same GO:0005576, no biological process — the only difference was prose about the substrate. The same change that introduced it had folded circulating plasma DNA into the first entry's description rather than giving plasma DNA its own entry, so the treatment was inconsistent. Surface DNA is now a substrate of the first core function, with its three evidence quotes moved there, and core_functions holds two genuinely distinct entries, extracellular and nuclear.
The proposed term's load-bearing claim is now a verified enumeration rather than an assertion. It previously said "nothing under DNA catabolic process (GO:0006308) occurs in the extracellular region". Checked against the pinned GO build, GO:0006308 has exactly three live is_a/part_of descendants — GO:0006309 apoptotic DNA fragmentation, GO:0032043 mitochondrial DNA catabolic process, GO:7770014 spermatid mitochondrial nucleoid elimination — and none is extracellular. The reviewer also asked about the endonucleolytic and exonucleolytic children; GO:0000737 and GO:0000738 are obsolete, which is why they are absent from the descendant list. That is recorded in the justification, with the query stated, so the next reader can re-run it instead of taking it on trust.
Also: the GO:0006308 row's reason had attributed two preprint-only results (the DNASE1 comparator, and the identity of the variant as R206C) to the published paper; they are now attributed to PMID:41280093. And the row now says why it stays ACCEPT rather than becoming a MODIFY onto the proposed term — the nuclear core function needs the general term, so replacing it here would leave that activity with no process.
Round 2 approved the review and raised three non-blocking points. Two are taken because they are not really style.
"Carriers" understated the cohort, and contradicted this file's own finding statement. Core function #1 said homozygotes' excess surface DNA belonged to "carriers of the poorly secreted R206C variant". The finding statement a few hundred lines above reads "Homozygotes for the secretion-defective R206C risk allele (rs35677470) have more surface DNA", and that is what the preprint measured. Note the word is not wrong everywhere: PMID:36708011's systemic-sclerosis result genuinely is about carriers digesting apoptotic-cell DNA in plasma. It was wrong for the surface-DNA result specifically. This is the third time in this review that prose has disagreed with a structured field in the same file, after the B-cell substrate claim and the preprint attribution; the pattern is that the careful version gets written into the record and the loose version into the summary.
The description had acquired a methodological clause. It read that "the mitochondrial origin and 8-oxo-dG enrichment of surface DNA were established in cell lines rather than sequenced from primary B cells" — which is how the evidence was obtained, not what the gene does, and the description-quality rule reserves that field for a project-independent biological summary. The fix is not to drop the caveat but to put it only where it belongs: core function #1 and the GO:0006308 reason both still carry it. The description now states the biology declaratively, that the tethered DNA is of mitochondrial origin and carries oxidative damage and that B-cell surface DNA is removed by the secreted enzyme and not by DNASE1. Those are the published abstract's own claims about select mammalian cells and about B cells respectively, so nothing is overstated by the shorter form.
The descendant check is now re-runnable rather than merely sourced. The justification said it had been checked against the pinned GO build; it now gives the query, runoak -i sqlite:go.db descendants -p i,p GO:0006308, and what it returns (four lines including the root). This was the reviewer's own stated residual: it corroborated the three terms from cache/enums/gobiologicalprocessenum_*.csv and proved that cache excludes obsoletes by using known-obsolete GO:0000002 and GO:0000003 as controls, but could not confirm completeness, because the enum cache carries term lists without parent/child edges. The query closes that, and -p i,p is what restricts the walk to is_a and part_of.
Round 3 re-approved and raised three non-blocking points. Two are fixed; the third was a correction to a PR comment of mine rather than to the file, and it was right.
The demonstrative re-linked what two rounds had separated. Fixing the method clause out of description left it reading that the tethered DNA is of mitochondrial origin and carries oxidative damage, and then "On B cells this surface DNA is removed by the secreted enzyme". The anaphor points back at the mitochondrial, oxidatively-damaged DNA, so a reader skimming the sentence still arrives at "B-cell surface DNA is oxidized mtDNA" — the round-1 blocker, restored by a pronoun while removing a clause. Dropping the one word closes it: "On B cells, surface DNA is removed by the secreted enzyme and not by DNASE1." The unscoped first clause stays because it is the paper's own framing, in the abstract and in its title.
This is the fourth instance in this review of the same shape, and the pattern is now clear enough to state as a rule for next time. Each time, the careful version was already in a structured field and the loose one went into prose: the B-cell substrate identity (reference_review right, description wrong), the preprint attribution (supported_by right, reason wrong), the cohort (findings[].statement right, core_functions wrong), and now the anaphor (the separation held in core function 1, lost in description). The file contained its own refutation every time, so the check that would have caught all four is to re-read the structured fields on the same claim before pushing prose about it — cheaper than a review round, and it does not depend on a reviewer noticing.
The query needed its portable form. The justification gave runoak -i sqlite:go.db ..., which resolves only with a go.db in the working directory; my own invocation had used an explicit path. It now reads runoak -i sqlite:obo:go descendants -p i,p GO:0006308, verified to resolve to the same build and return the same four lines. Reproducibility was the whole point of adding the query, so the short form defeated it.
The point I got wrong was in a comment, not the file. I wrote on the PR that both halves of the new description are the published abstract's own claims. The DNASE1 comparator is not: the published abstract says only that surface DNA is "cleared by the secreted nuclease DNASE1L3", and the DNASE1 comparison is preprint content. That is the same published-versus-preprint distinction corrected on the GO:0006308 row earlier in this PR, reappearing in my own prose about having corrected it. The file needs no change — description is uncited and PMID:41280093 is cited for that result elsewhere in the review — but the slip is worth recording, because it is the attribution half of the same pattern above.