BLOS1/GCN5L1 contributes structural work to two distinct trafficking assemblies. BLOC-1 supports endosomal cargo sorting and lysosome-related organelle biogenesis. Its reconstituted pallidin-Cappuccino-BLOS1 trimer and the depletion-dependent loss of BLOC-1 components support an integral scaffold role, rather than identifying BLOS1 as a free-standing motor or enzyme [PMID:15102850; PMID:22203680; PMID:16837549; PMID:25898167]. BORC shares BLOS1, BLOS2 and SNAPIN with BLOC-1 but contains a different set of other subunits. Human HeLa BLOS1 depletion destabilizes BORC and disrupts lysosomal Arl8b and peripheral lysosome distribution PMID:25898167. The two complexes are represented separately in the core functions; complexes are not placed in the location slot.
GCN5L1 also has cytosolic and mitochondrial pools. Human HepG2 mitochondrial localization and endogenous mouse subcompartment studies are distinct experiments. Its association with ATP5F1A and NDUFA9, the reduction in acetylation after knockdown, and respiratory counter-regulation support an acetylation-regulatory role PMID:22309213. Selected Results show increased respiratory activity following depletion, despite a differently worded section heading. The aerobic-respiration assertion is refined to negative regulation of cellular respiration in the tested context. No intrinsic acetyltransferase MF is asserted in the core. Native mitochondrial extract dependence does not disprove catalysis; it leaves the catalytic-site versus cofactor/assembly distinction unresolved. Human AC16 knockdown/overexpression experiments support GCN5L1-dependent Rictor acetylation and stability, while the authors leave catalytic-domain/null-mutant work open PMID:38281616. Source row 50 remains UNDECIDED, preserving the positive cellular findings.
All 56 source assertions and both alternative products are preserved in their original order and complete source fields. The 57 raw GOA rows retain a projected duplicate with differences in ECO ID, assigned source and date. No NEW annotation is proposed, so no process is manufactured from necessity, comparator absence or pathway completeness. Both isoforms P78537-1 and P78537-2 are retained; partner isoform identifiers in interaction assertions are also unchanged. None of these data establish that all reported activities are unique to one BLOS1 isoform.
Supported generic physical associations are retained as non-core, consistent with the user's action definitions. BLOC-1/BORC membership does not establish direct pairwise contact between every subunit. The exact HuRI records for HGS, MAGEB4, SNAPC5, DCTN2, USHBP1 and FBXO44 were not independently inspected, so those six source assertions remain UNDECIDED rather than removed. SNAPIN and pallidin associations have independent complex evidence. The extracellular tear-proteomics assignment is also UNDECIDED because the BLOC1S1-specific protein result was not accessed PMID:23580065.
The two IBA assertions are retained. No PAINT ancestral-node reassignment or target-specific evolutionary loss is proposed. Human P78537 appearing among experimental descendants for BLOC-1 membership is not circular, and the number of listed donors is not treated as evidence quality.
The neuronal assignments require careful separation of BLOC-1 cargo sorting from BORC lysosome transport. The official MGI Bloc1s1 donor graph links the two neuronal transport terms to PMID:21998198. Selected primary experiments concern PI4KIIalpha delivery from cell bodies to neurites in dysbindin-deficient mouse neurons and complementary rat PC12 cargo perturbations, with human neuronal-cell co-complex evidence. They support complex-level cargo sorting, not universal movement of every synaptic vesicle protein or a BLOS1-specific neuronal knockout. Neurite-development evidence in PMID:19546860 uses pallid mouse hippocampal cultures. These contextual assertions and the logically derived axon localization are retained as non-core. BLOC-1 early-endosome localization in PMID:16837549 was measured with pallidin/dysbindin antibodies; it is not represented as BLOS1-specific imaging.
The two NAS size assertions remain UNDECIDED. In the selected PMID:31314175 comparisons, Diaskedin/Myrlysin and Lyspersin perturbations are not equivalent for size even when positioning changes are shared. Complex membership does not settle BLOS1's size role, but no negative BLOS1 result is inferred. The independent bounded consultation agrees for source rows 48/51 and also for row 50; it read the actual term definitions and selected primary assays, with no whole-paper or all-figure claim.
All ten seeded PMID metadata/abstract blocks, the Reactome429815 summary and all 56 source rows were read. Additional already-cached primary references 16837549, 19546860 and 21998198 were read as described above. PMID15102850,22203680,22309213,23580065 and16837549 remain abstract-only in the canonical cache. Selected official indexed full-text passages supplemented 22203680 and 16837549. The staged full text of 22309213 was consulted separately without upgrading the canonical cache or quoting staged-only text as canonical supporting_text. Selected full Results/Discussion from 25898167, 31314175 and 38281616 were examined, but all images/supplements and all screen pair records were not. Existing shorter HTML-derived 25416956/33961781 caches and all other source preimages remain unchanged.
The normal auxiliary import added only 31314175 and 38281616 after independent source and copier review; all nine preexisting normal references and both family files were protected. GO term definitions were checked against official AmiGO: integral protein-complex scaffolding, the cytoplasmic lysosomal face, negative regulation of cellular respiration, BORC/BLOC-1 membership and the three disputed size/catalysis terms. Exact verbatim anchors are confined to the core support fields and independently checked against canonical caches; their aggregate per-source word totals are recorded in the proposal quote ledger. This notes entry adds no further direct quotations.
The official PubMed abstract of PMID:41887224 reports human BLOC1S1 knockout cell/iPSC-neuron and rescue evidence for lysosomal/autophagic and pigment-cell effects (DOI 10.1016/j.ajhg.2026.02.024; PMC13087471). Only its complete official abstract was read; direct PMC access returned a challenge. A single normal fetch-pmid attempt with writable UV cache/tool directories failed DNS (89d8a7 to b03e59, no output). This uncached paper is optional corroboration, not a formal review reference or a validated quotation. Reference batch 100 was reserved separately without a retry or a fabricated cache. The proposal stands on the available normal sources.
The single normal Falcon wrapper with its configured perplexity-lite fallback failed before either provider ran. The invocation omitted UV_TOOL_DIR, and uvx could not create a temporary file under the read-only default tool directory (d572a4). This was a local startup failure, not provider research or a dependency-resolution result. No provider retry or provider-named research file was authored. All scientific synthesis here is manual. Later authorized normal invocations use writable UV_TOOL_DIR as well as UV_CACHE_DIR.
This initial notes file is prepared in TMP for independent scientific review. Canonical application, validation history and publication remain separate steps. Later notes must be appended rather than rewriting this journal.
The normal schema, GOA, title and supporting-text validation passed on 2026-10-01 (761ae4 startup, a62b62 completion; detailed log09ddb3). Seventeen warnings remain: fifteen generic protein-binding rows retained as non-core under the user's action definitions, and two warnings that the scaffold core MF is not duplicated as a NEW annotation. Those warnings do not change the evidence-based decisions. All source-cache bytes and the canonical seed remain unchanged; independent whole-gene scientific review is still pending.
The additional normal validation with --terms also passed (b2c431 to d6115b), with the same 17 warnings and no errors. All formal reference caches were present, and their bytes were unchanged after validation.
ROOT independently reviewed all 56 annotation decisions and three core functions against selected primary Results and Methods, donor evidence, and official GO definitions (ccff85). The exact proposal was applied (cee211), then passed normal gene-path and term validation (c75370), HTML rendering (9cf7fc) and history validation (c23ee2). The 17 advisories comprise 15 supported contextual binding associations retained non-core under the supplied ActionEnum and two scaffold core terms not duplicated as NEW annotations. The rendered page embeds the reviewed YAML. All 56 source objects and both alternative products are preserved; 57 raw GOA rows contain one metadata-only duplicate after projection. Ten unresolved assertions remain explicit. Source caches are unchanged. No whole-paper, figure-image, supplement or repository-wide audit is claimed, and no successful provider report is claimed.
The lysosomal-location rationale now states the source distinction directly: the Reactome BLOC-1 assembly event supplies no lysosomal localization. The replacement location comes from the BORC study, PMID:25898167, which separates lysosome-associated BORC from BLOC-1. BLOS1 is shared between those complexes; that does not transfer each complex's location to the other.
All six HuRI pairs are present in the normal UniProt record with NbExp=3. Their reasons now explicitly acknowledge this positive corroboration. The pair-level primary confirmation and functional context remain uninspected, so the six source assertions remain UNDECIDED. The 15 independently supported complex associations remain KEEP_AS_NON_CORE under the supplied ActionEnum: lack of specificity alone does not establish biological error. No pair-specific scaffold activity is inferred merely from an interaction.
The two scaffold functions are already machine-readable GO:0140378 entries in core_functions, supported by BLOC-1 reconstitution and BORC assembly evidence; they are not prose-only conclusions. This bounded follow-up does not duplicate them as NEW annotations. The mitochondrial regulatory role remains separated from unresolved catalytic chemistry. Cellular acetylation changes do not isolate the acetyl-transfer step, but the IMP evidence code alone is not a reason to negate molecular-function evidence. No catalytic subunit or complex is invented. Tear-fluid detection also remains unresolved; lack of a conventional signal peptide does not disprove extracellular detection.
The clinical paper was reverified against PubMed PMID:41887224 and the publisher abstract. The PubMed record gives PMCID PMC13087471 and the matching DOI. The publisher abstract describes 11 affected individuals from seven families, BLOC1S1 knockout cellular and iPSC-neuron phenotypes, and variant-rescue results. This verifies the earlier citation; the normal local publication cache is still absent, and no new cached-reference entry or quote is added. Earlier source-fetch reservations in this journal are historical and do not establish a completed or pending import of this paper.
UniProt's statement about isoform 2 being the form found in non-primate orthologs is an inferred comparative statement. It does not identify which human isoform was assayed or establish that every human ISS/IBA source assertion is isoform-specific. Both existing alternative-product records and all original source objects remain unchanged. The previous failed provider attempt remains documented; this follow-up makes no provider-generated or complete-paper claim.
Totals remain 22 ACCEPT, 21 KEEP_AS_NON_CORE, 10 UNDECIDED and 3 MODIFY, with three core functions and no NEW annotations. Only eight review rationales and this appended journal entry change; all source files and cached publications remain unchanged.
Independent scientific review approved the eight rationale-only changes. Focused canonical validation passed with the 17 existing advisories: 15 generic-binding decisions and two scaffold-core terms not duplicated as NEW annotations. Rendering and new history validation passed. All 56 source assertions, existing actions, references, quotations, two alternative products and three core functions remain unchanged.
The current BLOC1S1 review retains 15 supported GO:0005515 protein binding assertions as KEEP_AS_NON_CORE and leaves six HuRI assertions UNDECIDED. The full review remains 22 ACCEPT, 21 KEEP_AS_NON_CORE, 10 UNDECIDED and 3 MODIFY across 56 source assertions, with three core functions and no NEW annotations. This section explains the retained assertions; it supersedes the earlier shorthand that described all 15 as complex associations. Thirteen concern BLOC-1 or BORC partners, and two concern the mitochondrial proteins ATP5F1A and NDUFA9.
The annotation-reviewer guidance and generic-binding validator normally favor a supported, more informative molecular-function replacement or removal of generic binding for lack of functional information. That policy explicitly allows removal without calling the interaction false. It applies to new reviews, and this review does not qualify for its legacy allowance. The 15 validation advisories therefore identify a real departure from that default; passing validation does not resolve the policy disagreement.
The task instruction for this review is to preserve supported generic interactions outside the core when no justified specific replacement is available, and not to remove them merely because the term is uninformative. That instruction governs these 15 decisions. Their retained information is the source-specific association and named supporting partner, while the core synthesis separately describes the biological work of BLOS1. This is a gene-specific application of the instruction, not an amendment of the general validator or skill.
| Retained source assertions | Positive evidence | What that evidence does not establish |
|---|---|---|
| Three SNAPIN/BLOS2/pallidin associations from PMID:15102850(https://pubmed.ncbi.nlm.nih.gov/15102850/) and three from PMID:22203680(https://pubmed.ncbi.nlm.nih.gov/22203680/) | Endogenous BLOC-1 co-isolation and recombinant eight-subunit assembly support the shared complex. BLOS1 and pallidin also occur in the reconstituted pallidin–Cappuccino–BLOS1 trimer. | The canonical caches are abstract-only; selected full-text assembly passages were read separately. Every original pair experiment was not independently reconstructed, and shared assembly does not show that every pair contacts directly. |
| One assertion with seven BORC partners from PMID:25898167(https://pubmed.ncbi.nlm.nih.gov/25898167/) | Affinity isolation, tagged-subunit co-isolation and recombinant assembly support BLOS1 in BORC. | This is one source assertion with seven supporting entities, not seven independent binding rows or seven isolated binary contacts. BORC and BLOC-1 are distinct assemblies. |
| One pallidin assertion from PMID:25416956(https://pubmed.ncbi.nlm.nih.gov/25416956/), two SNAPIN/pallidin assertions from PMID:32296183(https://pubmed.ncbi.nlm.nih.gov/32296183/), and three SNAPIN/BLOS2/pallidin assertions from PMID:33961781(https://pubmed.ncbi.nlm.nih.gov/33961781/) | The reported screen associations have independent BLOC-1 assembly context from the studies above. | Exact screen pair records were not independently reconstructed. Binary-screen reports and affinity-purification co-recovery are distinct evidence types; neither is promoted here into an isolated functional interface. |
| Two ATP5F1A/NDUFA9 assertions from PMID:22309213(https://pubmed.ncbi.nlm.nih.gov/22309213/) | The normal UniProt record and the previously inspected selected co-immunoprecipitation results support mitochondrial respiratory-protein associations. | These proteins are not BLOC-1/BORC subunits. The canonical cache is abstract-only, and selected full-text reading was separate. Association and cellular acetylation effects do not identify an intrinsic GCN5L1 acetyl-transfer catalytic site. |
The six HuRI partners HGS, MAGEB4, SNAPC5, DCTN2, USHBP1 and FBXO44 remain UNDECIDED. UniProt records each pair with three experiments, which is positive corroboration. The exact primary confirmation and biological context were not inspected, so these rows retain their existing uncertainty. Their treatment does not imply that every retained screen pair was independently verified at the same level.
BLOC-1 and BORC membership and their scaffold functions already appear in the review's core synthesis. A partner-specific association and membership in an assembly encode different relationships. Replacing each binding row with scaffold activity would additionally claim that its cited pair experiment establishes that activity; the available evidence does not support that substitution for every row. The mitochondrial associations likewise do not justify manufacturing a catalytic function. Generic binding remains outside the core, and none of the existing interaction decisions, reference quotations, product records or core functions changes in this documentation follow-up.