DOCK6 has one integrated GEF role connecting Rho-family activation to actin organization. The review retains ten imported source assertions and proposes no new annotation. The supported specific refinements are cytoplasm to cytosol, small-GTPase signaling to Rho signaling, and Rho regulation to positive regulation. The latter reflects the performed exchange/activation step, not developmental necessity. Positive generic association is retained as non-core under the scoped project instruction.
PMID:17196961(https://pubmed.ncbi.nlm.nih.gov/17196961/) reports Rac1/CDC42 exchange and explicitly identifies mouse N1E-115 cells. Its full Methods remained inaccessible, so the introduced construct species and purification cannot be reconstructed from the abstract. PMID:23462102(https://pubmed.ncbi.nlm.nih.gov/23462102/) describes mouse Dock6 phosphorylation/dephosphorylation during neuronal development; the complete abstract was read, not its full Methods. PMID:27693507(https://pubmed.ncbi.nlm.nih.gov/27693507/) describes compensatory human-cell responses after chronic DOCK6 loss, distinct from acute depletion. Human patient-fibroblast actin and morphology phenotypes in PMID:21820096(https://pubmed.ncbi.nlm.nih.gov/21820096/), Results and Figure 4, corroborate the biological role but are not a direct human exchange assay. Its mouse expression experiments remain separate.
The 2020 human-cell screen in PMID:32203420(https://doi.org/10.1038/s41556-020-0488-x) detected no RhoA/Rac1/CDC42 activation by full-length DOCK6 or its engineered DHR2 fragment under its tested conditions. Figure 2a source data give full-length mean ratios 1.025313, 0.995138 and 0.988886, respectively, below the stated thresholds 1.0619, 1.0121 and 1.016. The DHR2 fragment means are likewise below threshold. These conditional negatives are preserved alongside positive earlier studies; they do not establish universal inactivity. The paper discusses activation discrepancies and autoinhibition generally without demonstrating a specific DOCK6 release mechanism.
The official Q96HP0 record and FASTA identify one displayed 2047-residue product. Supplementary Table 1 row 78 identifies the human DOCK6 library construct; translating its coding sequence reproduces the current protein exactly. Tags and the 1586–2047 DHR2 fragment are engineered constructs, not natural products. Core substrate identifiers RAC1/P63000 and CDC42/P60953 are verified human identities; they do not transform the older neuronal study into a purified human-protein experiment.
The genuine original PDF, journal page 506/Figure 6f, states: “LRCH4 binds all C-Dock RhoGEFs, but not DOCK9.” The targeted HEK293T FLAG-LRCH4/mCherry-DOCK6 co-immunoprecipitation corroborates the supplementary AP-MS association. Two exact IntAct pair records were sequence-authenticated against Q96HP0 and O75427. The targeted record's Figure 6E locator differs from the original Figure 6f. Neither co-IP nor the spoke-expanded AP-MS record is represented as purified binary affinity. The DOCK7 DHR2 sufficiency experiment and DOCK8 ER recruitment are not assigned to DOCK6. Figure 6d supports DOCK6 redistribution toward the periphery with LRCH2.
HPA target data identify supported cytosol ICC/IF. The data and target identifiers were inspected, not the microscopy pixels. Reactome R-HSA-1011598 contains input set R-HSA-1012978 with explicit member R-HSA-1012966, human Q96HP0 residues 1–2047 in cytosol. This positive path corroborates localization. The event title's RhoGEFs wording differs from its Rho-GTPase substrate description. Perinuclear enrichment attributed to the inaccessible 2007 experiment remains UNDECIDED; diffuse cytosol does not reproduce or refute that specific distribution.
The current PTHR23317 tree contains 430 nodes. DOCK6's exact MSA sequence identifies PTN002511218 on a 21-node path. Literal IBDs place GEF and Rho regulation at PTN000593987, and cytosol at PTN002762118. Six current experimental donor joins were inspected; the plant seed remains in the literal IBD without a new independently resolved tree join. The target's own cytosol evidence is legitimate descendant grounding. PMID:22634526(https://pubmed.ncbi.nlm.nih.gov/22634526/) supports fly Zir genetic interactions with Rac2/Cdc42, not purified human catalysis. IPR026791 supplies the genuine DOCK-family mapping without importing DOCK1-specific interactions.
Current typed ancestry places GO:0007266 below GO:0007264 and GO:0035025 below GO:0035023. The positive-regulation term is not treated as an is_a/part_of child of Rho signaling itself. No unsupported new process, stable complex, adaptor activity or independent natural short product is asserted.
The 2007, 2013 and 2016 original Methods were not recovered after genuine publisher/repository attempts. The 2020 original and supplements were recovered independently while its protected publication cache remains abstract-only. Reference full_text_unavailable values describe the actual cached records. Biological findings are separated from access notes. New quotations total 13 words from PMID:17196961 in the YAML core and 8 words from PMID:32203420 above, each used once. The original-only Figure 6 quote is not placed in cache-validated YAML evidence. Genuine Falcon authentication failure and fallback quota failure are retained in the acquisition receipts; no provider report was invented. The source-evidence sidecar supplies stable navigation and exact assay/table locators.
The perinuclear annotation is retained as KEEP_AS_NON_CORE. The curated UniProt Q96HP0 record explicitly assigns ECO:0000269 and PMID:17196961(https://pubmed.ncbi.nlm.nih.gov/17196961/) to its cytoplasmic and perinuclear location statements. The earlier unresolved verdict gave excessive weight to unavailable original localization Methods despite this positive curated attribution. The exact assay and introduced-construct species remain unverified; the abstract's mouse N1E-115 context is not converted into an independently read human localization experiment. The valid SL-0198 mapping has no demonstrated propagation failure. The pool remains contextual because a distinct perinuclear mechanism is not established.
The broad cytoplasm refinement now preserves both cytosol and perinuclear region of cytoplasm. Both have cytoplasm ancestry, and neither is an ancestor of the other. These nonexclusive pools combine the curated location evidence with the supported HPA target ICC/IF data and the independently inspected human localization study PMID:32203420(https://doi.org/10.1038/s41556-020-0488-x). No new image-pixel interpretation is claimed. Cell-peripheral redistribution remains descriptive: proximity to the cell surface alone does not establish plasma-membrane binding or a named cortex compartment. Perinuclear cytoplasm is not nuclear-membrane lumen.
The core remains one GEF unit with Rac1 and CDC42 substrates and positive regulation of Rho signaling. Existing row 7 records the performed exchange step in a Rho signaling cassette, while row 8 records its positive regulatory effect. The current ontology connects GO:0035025 to GO:0007266 by positively_regulates, not is_a or part_of. Those terms are not declared ontologically redundant; the compact core uses the signed effect to summarize the same supported exchange mechanism rather than adding a second pathway label solely to mirror annotation rows. The original condition-specific negative human FRET screen remains explicit, alongside positive neuronal studies and human loss-of-function phenotypes.
No separate actin-assembly NEW assertion is introduced. Human patient-fibroblast actin and morphology defects in PMID:21820096(https://pmc.ncbi.nlm.nih.gov/articles/PMC3155174/) establish a consequence of DOCK6 loss, not an additional directly assayed actin-assembly step performed by DOCK6. This does not deny the supported role of GEF signaling in cytoskeletal regulation. The top biological description is shortened; the detailed organism, construct and assay boundaries remain in the annotation reasons, core and reference assessments.
Focused cache-exact support is added without repeating quotations in these notes. Newly reproduced word counts are 13 from the UniProt record, nine from PMID:17196961 and ten from PMID:21820096. The inherited thirteen-word PMID:17196961 core anchor is retained once, for 22 words from that source in the effective YAML. The older eight-word PMID:32203420 original-only note excerpt is unchanged. The UniProt FUNCTION block's mixed ECO:0000250/ECO:0000269 and similarity-qualified neurite statement remain explicit in the new reference assessment. All ten imported source objects, the single 2047-residue product, core terms and substrates, caches and earlier notes are preserved; only the perinuclear action changes.