The genuine normal intake supplied 40 assertions, no alternative-products field and 15 reference identities. All source fields remain exact. The review resolves 26 assertions as ACCEPT, 7 as KEEP_AS_NON_CORE, 6 as MODIFY and one as UNDECIDED. It adds no annotation. One core describes CLDN16 contribution to the assembled paracellular tight-junction channel, with renal reabsorption and the between-cell transport route.
The important distinction is the route taken by ions. GO:0015095 and GO:1903830 describe passage across a membrane, whereas CLDN16 contributes to a junctional pathway between epithelial cells. The original PMID:10390358 abstract supports renal tight-junction and magnesium-resorption biology, but its full text was not recovered after publisher failures and bounded searches. No original assay species, figure, or absent experiment is inferred from that access limit. The intracellular-cation homeostasis assertion therefore remains UNDECIDED.
Later direct evidence preserves the magnesium result. Human CLDN16 expressed in MDCK-C7 layers increases magnesium permeability measured by atomic absorption with Ussing-chamber methods; some variants reach junctions but fail to increase flux (PMID:16528408(https://pubmed.ncbi.nlm.nih.gov/16528408/)). Human CLDN16 and CLDN19 in LLC-PK1 layers yield strong sodium/chloride selectivity, with a different magnesium response in that epithelial background (PMID:18188451(https://pubmed.ncbi.nlm.nih.gov/18188451/)). Dog CLDN16/19 reconstitution in claudin-deficient MDCKII cells produces substantial magnesium and calcium permeability, with endogenous CLDN16 explicitly controlled and discussed (PMID:39821681(https://pubmed.ncbi.nlm.nih.gov/39821681/)). These are complementary, context-dependent experiments, not grounds for declaring magnesium permeability absent or converting a junctional channel into a cellular magnesium pump.
The broad intercellular-transport and metal-ion-transport assertions are refined to the existing paracellular process. The latter is biologically valid as written: its definition includes movement between cells. The replacement makes the route explicit and changes the axis of specificity; it does not claim an ontology parent/child relationship or discard metal-ion selectivity. No NEW process is needed because both the channel and process already occur in the seed.
The current UniProt record describes a 235-residue protein and supplies no alternative-products object. The original human studies distinguish a second-start construct from a historical 305-residue form. The 2003 study uses the short human construct, and the recovered 2005 Methods/Results explicitly compare start sites. Older residue numbers and transcript identifiers are retained at their reported scope rather than silently converted into current product identities (PMID:14628289(https://pubmed.ncbi.nlm.nih.gov/14628289/), PMID:16234325(https://pubmed.ncbi.nlm.nih.gov/16234325/)). The exact normal abstract for the latter remains unchanged; selected original indexed full text supplies the external assay access.
Human CLDN16 associates with CLDN19, and CLDN19 supports its incorporation into junctional strands. CLDN16 alone does not form autonomous strands in the tested fibroblast or claudin-deficient epithelial systems. Cellular complex, interface-perturbation and permeability experiments support a joint function, without determining an invariant native stoichiometry or atomic junctional structure (PMID:18188451(https://pubmed.ncbi.nlm.nih.gov/18188451/), PMID:26446843(https://pubmed.ncbi.nlm.nih.gov/26446843/)). Mouse depletion and separately assayed interaction panels support reciprocal dependence of junctional localization (PMID:19706394(https://pubmed.ncbi.nlm.nih.gov/19706394/)); selected original PMC full text was read while the genuine normal abstract was preserved.
The 2017 study specifies tagged human CLDN16 and CLDN19 transcripts and reports cis FRET, including weaker positive CLDN16 self-association. Its native mouse/rat tissue studies remain separate. Positive cis FRET does not establish autonomous strands or trans adhesion, and it is not erased by the older negative homomeric two-hybrid result (PMID:28028216(https://pubmed.ncbi.nlm.nih.gov/28028216/)). The original Q8N6F1-2 partner identifiers are preserved without claiming a new exact sequence audit of every construct.
The 2012 paper explicitly pairs mouse CLDN14 AF314089 with human CLDN16 AF152101 and CLDN19 BC030524. Two-hybrid and reciprocal coimmunoprecipitation results support the CLDN14–CLDN16 association; three literal MINT records match this source pair (PMID:22373575(https://pubmed.ncbi.nlm.nih.gov/22373575/)). The newer model shows CLDN14 replacing CLDN16 in CLDN19-containing meshworks and changes in mouse renal junctions. Its inspected construct section does not explicitly settle every transgene species or accession. Host-cell identity is not used to infer protein species, and the new strand model does not negate the older pairwise association assays (PMID:41325524(https://pubmed.ncbi.nlm.nih.gov/41325524/)).
The CLDN16 terminal peptide binds ZO-1 PDZ domains with mutant and competition controls. Human HA-CLDN16 also associates with endogenous canine ZO-1 in MDCK cells. The generic O97758 binding assertion is refined to PDZ domain binding, already represented separately in the seed. This identifies the demonstrated binding class without transferring ZO-1 scaffold activity to CLDN16 (PMID:14628289(https://pubmed.ncbi.nlm.nih.gov/14628289/)). Other supported generic claudin associations remain non-core under the campaign policy; insufficient specificity alone is not a reason to remove a positive interaction. Raw interaction responses and all 42 fields remain in the temporary audit packet, while the published index retains the used literal projection and exact source joins.
O88551 is mouse Cldn1, not the mouse Cldn16 ortholog. Its experimental junctional, calcium-independent adhesion and self-binding annotations provide a real basis for the curated family transfers. The adhesion assertion is retained as non-core with the absence of an equivalent target-specific calcium-dependence assay explicit. CLDN16 partner-dependent junctional participation is not described as autonomous adhesion. Q925N4 and Q91Y55 are the mouse and rat Cldn16 orthologs with experimental transport and localization grounding.
PAINT node assertions are retained with term-specific experimental corroboration. Current descendant records and target experiments are authenticated; the historical tree, alignment and ancestral placement were not reconstructed. Target self-inclusion is legitimate, and donor count is not used as an evidence-quality score. Current ARBA00027156 and ARBA00088093 conditions match the target FunFam for tight-junction localization and paracellular transport. The historical matching route for the broad plasma-membrane ARBA component is not claimed reconstructed; direct target localization and the subcellular vocabulary mapping support that compartment.
All genuine normal source bytes and variants are preserved. Selected primary Methods, Results and Discussion were read with exact paragraph or indexed-line locators. No figure pixels or unread supplements are claimed in this review. The original PMID:10390358 remains abstract-only; PMID:16234325 and PMID:19706394 have externally recovered selected full-text evidence while their normal abstract caches remain unchanged. The compact source index provides identifiers, URLs, hashes, used literal data and paraphrased scope; it is not an independent experimental citation and contains no repeated quote inventory.
The genuine default Falcon attempt timed out after 90 seconds. Its actual Perplexity-lite fallback returned HTTP 401 quota failure. Manual primary research supplied the evidence synthesis, and no provider-labelled research text was fabricated. The authenticated GO-CAM index has no target match; no new process assertion is proposed. Initial ownership was checked against main d6a01fd305ee7d7fdc6fbc2d1c7a72eff02c8835, including a complete Git inventory of the unusually large PR #2049. Final applicability is checked separately against the current main and complete open-PR inventories, preserving this intake snapshot. Project completion requires publication, review and merge.
This entry supersedes the earlier replacement scope where magnesium transport was represented only by a general paracellular process. Current GO:0015693 includes magnesium movement between cells, while GO:0160184 specifies the route. Both now refine the existing metal-ion and magnesium-transmembrane assertions without creating a new annotation. Actual human CLDN16 MDCK-C7 flux experiments in PMID:16528408 retain the ion-specific evidence; dog CLDN16/19 reconstitution in PMID:39821681 remains explicitly nonhuman corroboration. The contributed channel MF stays GO:0160187. Its current ontology record has no child term; no identifier or new annotation is invented for divalent selectivity.
Actual PMID:26446843 Results (sucrose-gradient separation and chemical crosslinking, Figure 1) and Methods establish human expressed CLDN16/19 heterodimers in HEK293 and Sf9 systems. The gene-level positive pair is refined to GO:0046982, with native pore stoichiometry and modern isoform limits retained. The other exact CLDN19 source/product assertions remain positive and non-core.
The calcium-independent adhesion ISS assertion remains non-core. PMID:18188451 strand incorporation and PMID:28028216 cooperative contact enrichment distinguish partner-dependent structural participation from the negative standalone CLDN16 assay. They do not constitute a target adhesion assay with calcium removal; that specificity remains the preserved Cldn1-based family inference. The intracellular cation-homeostasis assertion remains undecided because the original source's precise intracellular readout is still unavailable; later transepithelial transport data neither establish nor refute that distinct assertion.
The external full-text access statements for PMID:16234325 and PMID:19706394 now explicitly distinguish actual inspected external passages from unchanged abstract-only repository caches. New short primary anchors are confined to YAML; no literal anchors are repeated here or in the source index. All original source assertions, reference identity pairs, selected cache bytes, prior notes and history are preserved.