DNAJB6 suppresses aggregation of client proteins, including polyglutamine-expanded reporters. The historically studied short b product acts in cytoplasmic and nuclear pools; the long a construct can suppress nuclear-targeted polyglutamine despite little activity against a cytoplasmic reporter. These substrate/compartment distinctions are important: the long product is not generally inactive, and a single luciferase-refolding result is not a universal test of all clients. PMID:20159555(https://pubmed.ncbi.nlm.nih.gov/20159555/), PMID:21231916(https://pubmed.ncbi.nlm.nih.gov/21231916/), PMID:22366786(https://pubmed.ncbi.nlm.nih.gov/22366786/)
Purified DNAJB6b can suppress polyglutamine aggregation without added HSP70. Its J-domain nevertheless has a distinct, regulated HSP70-binding and ATPase-activating capability. In the 2023 experiments, the exposed J-domain activates HSP70, whereas the WT JD-GF construct and full-length WT are autoinhibited under the tested conditions. Engineered disease variants can release that inhibition. The physiological factor or client that releases normal GF inhibition remains unresolved; client transfer is a proposal, not a demonstrated delivery mechanism. PMID:37923706(https://pmc.ncbi.nlm.nih.gov/articles/PMC10624832/), Results and Figures 2, 4 and 6; Methods, protein constructs and ATPase assay.
The current definitions of protein folding and protein-folding chaperone require productive folding, while unfolded-protein holdase activity additionally describes maintaining a client for delivery. Antiaggregation alone does not settle these terms. The unresolved folding and obsolete unfolded-protein-binding assertions therefore retain their positive source evidence without a forced folding/carrier replacement. The antiaggregation core has a biological-process term and description but no invented molecular function. The actual PAINT placement is valid and positively reconstructed; this uncertainty is not a claim of incorrect ancestry or universal loss of an ancestral activity. GO:0006457, GO:0044183, GO:0140309, PMID:23612975(https://pmc.ncbi.nlm.nih.gov/articles/PMC3682527/)
The keratin function is separately supported: human Mrj binds K18 and Hsp/c70 through distinct regions, colocalizes with K8/18 in HeLa cells and contributes to the organization of that filament network. The original complete abstract supports this conclusion, while its full Methods remain unavailable. PMID:10954706(https://pubmed.ncbi.nlm.nih.gov/10954706/)
Human HPA data report supported nucleoplasm and cytosol localization, without resolving individual products. The perinuclear annotation and human muscle Z-disc observations represent additional pools. The sperm-nuclear HDA annotation is retained broadly because the target table and its exact subnuclear pool were not independently recovered; somatic nucleoplasm evidence is not substituted for an unmeasured sperm subfraction. YTS membrane association remains undecided because the target-specific table is unavailable. A soluble sequence lacking a transmembrane helix does not contradict the membrane term, which includes attached proteins. HPA DNAJB6 subcellular data, PMID:10954706(https://pubmed.ncbi.nlm.nih.gov/10954706/), PMID:22366786(https://pmc.ncbi.nlm.nih.gov/articles/PMC3315599/), PMID:21630459(https://pubmed.ncbi.nlm.nih.gov/21630459/), PMID:19946888(https://pubmed.ncbi.nlm.nih.gov/19946888/)
Four current official products were reconstructed and checked against their own FASTAs: A, 326 residues; B, 241; C, 325; and D, 277. Product C has the official synonym a, so a historical lowercase label alone does not establish the modern accession. Product D replaces much of the canonical J-domain-containing N terminus; this sequence difference does not prove total inactivity. The well-studied b constructs and engineered disease variants are not silently assigned to every product. UniProt O75190, PMID:20159555(https://pubmed.ncbi.nlm.nih.gov/20159555/), PMID:37923706(https://pubmed.ncbi.nlm.nih.gov/37923706/)
The BRMS1, MLF1/MLF2 and other screening interactions are positive evidence. Source-specific pairs were traced through the actual assays, sequence ROGIDs and recorded fragments; generic partner association is retained as non-core when no additional activity is established. The targeted BAG3/HSPB8/STUB1 experiments are not automatically a proof of direct binary recognition of each component or stable CASA-complex membership. PMID:16919237(https://pmc.ncbi.nlm.nih.gov/articles/PMC1557677/), PMID:23414517(https://pubmed.ncbi.nlm.nih.gov/23414517/), PMID:25036637(https://pubmed.ncbi.nlm.nih.gov/25036637/), PMID:32814053(https://pubmed.ncbi.nlm.nih.gov/32814053/), PMID:33961781(https://pmc.ncbi.nlm.nih.gov/articles/PMC8165030/), PMID:40205054(https://pubmed.ncbi.nlm.nih.gov/40205054/), PMID:22366786(https://pubmed.ncbi.nlm.nih.gov/22366786/)
The mouse-derived DNA-binding annotation has a real experimental source. Mrj is recruited to an NFAT-associated TNF promoter complex, but the inspected ChIP and NFAT EMSA experiments do not resolve direct Mrj–DNA contact. This is retained as uncertainty, not rejected as erroneous ortholog transfer. The mutant-Parkin relocation result is likewise preserved at its actual scope; stronger clearance experiments on HSJ1a are not transferred to DNAJB6. PMID:16260608(https://pmc.ncbi.nlm.nih.gov/articles/PMC1280278/), PMID:20889486(https://pubmed.ncbi.nlm.nih.gov/20889486/)
The 2010 original was read as the authentic published PDF from the Groningen repository, especially Results, the comparison of nuclear versus cytoplasmic polyglutamine reporters and Figures 1–2. The nuclear a result should not be replaced by the broader UniProt shorthand that A does not inhibit huntingtin aggregation. Detailed selective HDAC4/acetylation mechanisms mostly concern DNAJB8; broad TSA effects include DNAJB6b. PMID:20159555(https://pubmed.ncbi.nlm.nih.gov/20159555/)
The 2011 original was independently recovered from the HAL accepted manuscript. Its DNAJB6b Results and figure labels were read through OCR with visual confirmation. The target-specific cellular antiaggregation/refolding comparison is distinct from the paper's purified DNAJB1/HSPA1A/HSPA6 experiments. These external original reads do not alter the abstract-only publication caches. PMID:21231916(https://pubmed.ncbi.nlm.nih.gov/21231916/)
The genuine publisher/full-text caches and failed access responses were preserved. An older purported 19946888 full HTML was actually a challenge page and supplies no original evidence. Full-text availability flags describe the committed caches; separately accessible originals and selected passages are documented in the source guide. The inherited Falcon report is genuine and unchanged, but primary papers and official records determine the manual judgments.
All 47 inherited source objects are retained. The current 61 raw GOA records normalize to 44 unique source triplets; the only three historical-only triplets use obsolete GO:0051082, and they remain in the review. No NEW annotation is added. Positive generic-binding NC decisions follow the scoped project instruction. This instruction does not verify an uninspected interaction or permit inferring a more specific activity from association alone. All quoted source anchors occur once in the YAML and remain within 25 words per publication; this notes file adds no verbatim publication excerpts.
The antiaggregation core now uses the repository's explicit proposed-term route: holdase chaperone activity, linked to a top-level proposal and NTR replacements on the three obsolete unfolded-protein-binding rows and the folding-chaperone IBA. This resolves the representation gap without claiming productive client refolding or the delivery step required by the carrier-holdase term. It follows the unfolded-protein-binding project and supersedes the earlier choice to omit a molecular-function representation. The proposed name is not an assigned GO identifier. No change to another gene is needed to represent DNAJB6's own supported activity.
The second core records HSP70 binding. The exposed J-domain and pathogenic-mutant ATPase activation remain positive, conditional non-core evidence; intact WT autoinhibition and the unknown physiological release mechanism are explicit PMID:37923706(https://pmc.ncbi.nlm.nih.gov/articles/PMC10624832/). The previous negative-result quotation on the ATPase row is replaced by a short anchor to the positive exposed-domain result. The older PMID:11896048 IDA is preserved without inventing access to its full experiment. Z-disc localization is included with the antiaggregation core, supported by the muscle localization and aggregate-pathology evidence PMID:22366786(https://pubmed.ncbi.nlm.nih.gov/22366786/). PMID:28233300 supplies complementary cellular antiaggregation evidence for the p.Pro96Leu variant; its dominant-negative filter-trap result is not a direct HSP70-interaction or ATPase assay and need not match the purified conditions or different variants tested in 2023.
The electronic DNA-binding transfer is marked over-annotated: the observed NFAT/HDAC-associated promoter recruitment remains positive, whereas direct Mrj–DNA recognition was not established by those experiments PMID:16260608(https://pubmed.ncbi.nlm.nih.gov/16260608/). Figure 6B tests whether Mrj blocks NFAT–DNA binding, not a purified Mrj–DNA interaction. This is a scope judgment about the electronic transfer, not a wrong-gene or nonbinding claim.
The nervous-system-process IDA remains UNDECIDED. Its full original is still inaccessible; brain enrichment and cellular-model results in the abstract do not establish which additional experiments the curator saw. The same access limit applies to the source-specific folding IDA. Neither is removed or declared unsupported solely from abstract silence. Each newly placed quote occurs once, within the aggregate per-source word limit. All 47 source dictionaries, four natural products, reference identities and machine-source files remain unchanged.