ATP13A2 is an ATP-dependent P5B polyamine transporter of the late endolysosomal membrane system. The principal function is export of polyamines, especially spermine, toward the cytosol. The review retains the original 120 annotation source objects and all three alternative products (Q9NQ11-1, Q9NQ11-2 and Q9NQ11-3). It proposes one transport core and no new annotations.
The current draft contains 39 ACCEPT, 70 KEEP_AS_NON_CORE, seven UNDECIDED, two REMOVE and two MODIFY decisions. Row numbers below are one-based in the unchanged seed order. The four monoatomic-transport rows retain non-core status after the recovered source assessment described below; all 120 actions remain unchanged.
PMID:31996848, ATP13A2 deficiency disrupts lysosomal polyamine export supplies the principal transport evidence. Its complete normal abstract was read. Selected original external Methods/Results establish an important distinction: purified-protein ATPase stimulation and flux in reconstituted yeast membrane-protein mixtures are separate assays. Human variant 2 was expressed in the yeast experiments. ATP-inside liposomes supported spermine uptake in the reported orientation, corresponding to endolysosomal export in cells. This is not described as homogeneous purified-protein transport, nor as an isoform-exclusive function. The abstract-only normal cache retains full_text_unavailable: true even though selected external original text was read.
The later PMID:37080960 study reports potassium-dependent ATPase measurements in human ATP13A2-expressing bulk membrane fractions, WT/D513N and knockout/rescue comparisons, permeabilized-cell 86Rb efflux and SH-SY5Y perturbations. The recovered normal XML record was read directly from the verified archive: the root reviewer read the full header/abstract, first complete Results including text legends, complete Discussion and complete Methods. The author independently reread the header/abstract, initial Results through the cation-transport section and complete Methods. Figure pixels, supplements and source-data spreadsheets were not inspected. XML repetitions are not additional observations. The full-text flag accurately describes the recovered record; it does not imply that every body section was inspected.
Rows 63/64/79/103 retain the reported potassium-associated activity as non-core. The experiments do not measure simultaneous purified H+/K+ flux or a coupling stoichiometry; the authors' proton-countertransport interpretation remains distinguished from the directly measured membrane ATPase, tracer and pH readouts. The historical Reactome import direction is not endorsed. The principal polyamine-export core remains unchanged, with no additional proton-pump core or NEW process. The one short Results quotation identifies the actual bulk membrane preparation.
A proposed experiment would use matched reconstituted wild-type and catalytic-control preparations and measure polyamine, potassium and proton flux separately under defined gradients, with ATP/no-ATP and orientation controls. This would test intrinsic coupling against cellular effects; no such matched experiment is asserted to have been performed.
PMID:23205587 reports human ATP13A2 expressed in CHO cells and polyamine uptake. Whole-cell accumulation in that assay does not reverse the mechanistic lysosome-to-cytosol direction established later. Existing polyamine and spermine terms already cover the activity; there is no new process assertion and no extra spermine child process in the core alongside the broader polyamine process.
GO:0015417 explicitly specifies an ABC-type transporter, with an ABC-type parent. ATP13A2 is a P-type ATPase. The two ABC-class annotations (51/52) are therefore removed for a concrete mechanistic-class mismatch, while the same primary paper's polyamine annotation (50) remains accepted. Broad reaction chemistry and older synonyms do not establish an ABC mechanism. GO:0015662, in contrast, refers to P-type transport of ions and is not restricted to monoatomic ions; charged polyamines do not contradict it.
Selected normal PMID:26134396 topology, targeting and lipid-binding Results were read directly, with separately documented construct/purification Methods. Human N-terminal residues1–187 were tested with GST controls, lipid overlays, charged-site mutants and liposome floatation. PA and PI(3,5)P2 binding support regulation of the transporter, not phospholipid flipping. These two existing binding functions are retained non-core; they do not require a separate molecular-function core. The article's optional five-residue insert and historical numbering were not mechanically mapped to a different modern isoform.
The membrane-spanning topology supports the core lysosomal and late-endosomal membrane locations. Broader vesicle and neuronal compartments are retained as context. For individually unread original localization images or proteomic tables, the reasons explicitly defer to the source curator where the target location is independently established. In particular, the PMID:17897319 abstract describes the proteomic survey without naming ATP13A2; no target-table verification is claimed.
The official Reactome R-HSA-5692480 event explicitly displays its ATP13A2 catalyst entity in lysosomal lumen. The separate entity page could not be opened in this reading. This is a documented source-compartment problem, rather than an inference from the event's transported cation. Row87 is modified to the existing lysosomal-membrane term using direct target topology. The historical probable cation-import model is not silently rewritten, and both immutable Reactome caches remain unchanged.
The original indexed PMID:22645275 Results explicitly name the seven co-IP partners corresponding to rows3–9 in HEK293 cells. V5-tagged ATP13A2 and the candidates were coexpressed and immunoprecipitated from whole-cell lysate. Complete Methods, supplements and figure pixels were not read; direct PMC rendering returned a challenge. These are verified cellular associations, not purified binary interfaces or automatically inferred adapter/enzyme activities. All seven remain non-core. The later PMID:27278822 blotting result did not confirm an obvious SYT11 association in that context; this qualifies rather than erases the earlier co-IP.
Selected normal PMID:27278822 Results, IP/MS Methods and Discussion support association with MYCBP2/TSC2 and TFEB/SYT11 regulation. TFEB, rather than ATP13A2, is the promoter-binding protein. No ATP13A2 DNA-binding or direct ubiquitin-ligase-inhibitor activity is proposed. MYCBP2 association is retained non-core alongside the existing regulatory process annotations.
Selected normal PMID:30538141 Results show human cellular co-IP of endogenous HDAC6 with overexpressed ATP13A2, together with tubacin-washout and rescue experiments. Endogenous ATP13A2 was not detectable with the available antibody. Mouse fractionation/MEF and fly experiments remain distinguished from human evidence. HDAC6 recruitment and fusion-pathway involvement are retained without claiming purified direct binding, intrinsic membrane-fusion catalysis or direct HDAC6 enzymatic activation.
The user-defined REMOVE action concerns a function unlikely to be correct. It is therefore not used solely because a generic binding annotation is uninformative. These eight supported associations remain non-core, with no manufactured replacement MF. This is an explicit difference from the skill's general recommendation to remove uninformative generic binding.
Human patient-cell and separately mouse-neuron zinc studies support homeostatic involvement without identifying every altered metal as an intrinsic transported substrate. Reported stress protection, autophagic trafficking, secretion and neuronal phenotypes are retained as secondary functions when supported. Context-dependent negative studies do not establish universal absence of lysosomal transport or erase positive results from different models.
Five ISS metal-binding rows explicitly use the yeast Q12697 donor. The cached abstracts and separately documented original 2013 Methods/Mn Results concern synthetic capped YPK9 peptides. This donor species is expected for ISS, not a misattribution. Conservation and function of the assayed sites in intact human ATP13A2 remain unestablished. A simple sequence-string absence is not an alignment or proof of loss. These five rows remain UNDECIDED.
Normal PMID:22285144 Methods, Results and Discussion describe a cross-sectional human SNP-by-environment study. The associated intronic variants have unresolved functional effects; no cellular perturbation assay was identified in the read sections. Row97 remains UNDECIDED rather than being reinterpreted as a demonstrated cellular mechanism. Row98 also remains UNDECIDED because the complete ATP13A2-specific discussion of PMID:23628791 was not accessed. An abstract foregrounding LRRK2 is not evidence of wrong-gene curation.
The complete normal PMID:22885599 commentary reports that ATP13A2 loss impairs proteolysis and wild-type restoration rescues lysosomal deficits. GO:1905166 instead specifies negative regulation of that process. Row120 is corrected to the already represented direction-neutral GO:1905165. This is a sign correction, not a new process or protease assignment. The existing macroautophagy and CMA regulatory assertions remain non-core, without claiming a CMA-selective translocation experiment. The source is a commentary summarizing prior experiments, not an independent primary assay.
Selected PMID:25548531 discussion explicitly connects ATP13A2 to impaired mitochondrial degradation and prior mitophagy studies, while leaving mechanisms unclear. Its contextual TAS annotation is retained; HDAC6-related speculation is not treated as an intrinsic ATP13A2 enzymatic activity.
The original 28 normal bibliographic records and 27 available abstracts were read; the additional PMID:37080960 record and abstract bring these totals to 29 records and 28 abstracts, along with the two normal Reactome summaries. PMID:26392192 has no abstract but its complete commentary body was read. Original source titles are taken from the normal frontmatter. full_text_unavailable tracks the normal cache flag and does not mean that every body section was inspected. Manual relevance/correctness judgments separately record target-level limits; UNVERIFIED for an unread target assertion does not imply that the PMID is invalid or retracted.
IBA assertions are treated as ancestral PAINT judgments. Neither donor count nor the human target's presence among descendant evidence is a defect. Exact PAINT trees were not reconstructed. No NEW annotation was proposed, and no external text was made into a fabricated publication cache. The normal provider attempt returned no report; these are manual notes, not provider output.
The final draft binds the canonical PMID:37080960 record byte-for-byte to the verified 135,791-byte recovery member (SHA256 184aba2e7aceb48fed4c2f5fe7533f8d4c4f5f7d070d97a1587147f993c43d2d). The source was imported by its separate recovery owner. This authoring preparation changes no canonical gene or source file.
The earlier neutral replacement for row120 is refined to GO:1905167 positive regulation of lysosomal protein catabolic process. The complete normal PMID:22885599 commentary describes defective degradation in ATP13A2-mutant patient fibroblasts and depleted human BE-M17 cells, and rescue with wild-type ATP13A2. This favors the positive direction while retaining a secondary homeostatic interpretation. ATP13A2 is not assigned protease activity or another core. TAS and the original source assertion remain unchanged: this short commentary summarizes earlier work and has no independent Methods section. No complete read of its underlying primary experiment is claimed here.
The three pre-existing neutral regulation rows remain supported broader assertions. Different evidence rows may legitimately overlap; eliminating duplication is not the rationale for a more specific replacement. This is a refinement of an existing MODIFY, not a NEW process assertion.
The transport core now uses its transport-specific PMID:31996848 anchor alone. The PMID:26134396 topology quotation remains represented in its reference finding and the location-related annotations, rather than serving as if it measured polyamine movement. Three short, cache-verbatim reference findings record the transport, topology and lysosomal-rescue evidence. Normal availability flags are unchanged; the 2020 transport cache remains abstract-only. Previously documented human variant-2, yeast-membrane reconstitution and purified ATPase assay boundaries remain in force, without a new full-paper read claim.
The eight verified generic associations remain non-core under the user's ActionEnum. No functional MF is invented from association alone. The five synthetic yeast-peptide metal-binding transfers remain UNDECIDED: a small alignment alone would not establish activity of those sites in intact human ATP13A2, and no new alignment or assay is claimed.
The applied follow-up passed focused validation (nine existing advisories: eight generic-binding decisions and the source-specific manganese-response action difference), rendering, and validation of the newly scaffolded history record. These checks do not establish a new repository-wide validation pass. The independent review covered the finite changes above rather than repeating the entire 120-row scientific audit.