Human FAM111A binds single-stranded DNA and associates with replication-associated chromatin. Its PIP box binds PCNA, and catalytic, DNA-binding and PCNA-interaction regions contribute to traversal of protein obstacles at replication forks. These are distinct tested roles: PCNA contact is not clamp-loader activity, and FAM111A is not a DNA polymerase.
Alabert et al., PMID:24561620, Results on nascent chromatin capture and novel replication factors, Figures 6–7, and Online Methods, combines human HeLa S3 nascent-chromatin proteomics with U-2-OS localization and TIG-3/U-2-OS RNAi. FAM111A is enriched on newly replicated chromatin, has Triton-resistant nuclear foci, and colocalizes with PCNA. In-vitro-translated FAM111A binds recombinant GST-PCNA; Y24A/F25A substitutions impair this PIP-dependent interaction. The named tagged OriGene construct is historical reagent provenance, not an independently verified match to every current transcript.
Depletion in that study reduces EdU incorporation and chromatin PCNA and delays S-phase entry. In contrast, the later HAP1 knockout study finds largely unchanged untreated fork speed and chromatin PCNA. The shared conclusion is replication-associated participation, especially at protein obstacles; universal PCNA loading or indispensable basal replication is not inferred. The direct PCNA interaction remains a supported non-core binding annotation after a targeted search found no demonstrated clamp-loading/unloading activity to substitute.
Kojima et al., PMID:32165630, Figures 2–6 and Methods, tests human FAM111A expressed in human cells, Sf21 cells or bacteria. Full-length and fragment assays are distinguished. EMSA, antibody supershift and fluorescence polarization support preferential ssDNA binding; the central 176–282 region and F231 contribute. This does not imply exclusive ssDNA recognition.
Figure 3 measures active-site-dependent cellular cleavage and trans-cleavage between differently tagged FAM111A molecules. Edman sequencing of an Sf21-produced R569H fragment maps an internal cut after Phe334, with cleavage-site substitution controls. These fragments are processed experimental species, not alternative transcript products. The paper explicitly reports failure to obtain protease activity from its purified recombinant preparation in vitro.
Hoffmann et al., PMID:32776417, the Results section “Human FAM111A and FAM111B are active proteases,” Figure 1G and Figure EV1C/D/F, and recombinant-protein purification Methods, supplies independent positive purified activity. Full-length human FLAG-FAM111A purified from yeast JEL-1 undergoes autocleavage, controlled by the catalytic D439N substitution and AEBSF. Human FAM111B is assayed separately. This positive preparation complements the earlier failed preparation; it does not retrospectively change that result. Neither source reconstitutes cleavage of a defined TOP1–DNA or PARP1–DNA substrate by purified FAM111A.
Human knockout/rescue, TOP1 covalent-complex staining and DNA-combing experiments support FAM111A's existing protein–DNA cross-link repair and fork-processing annotations. Catalytic-serine, PIP-box and DNA-binding mutants fail relevant rescue. TOP1 covalent complexes are distinct from PARP inhibitor traps, which are generally noncovalent; PARP1 knockout helps separate trapping from loss of PARP catalytic signaling. The hydroxyurea result does not establish a general fork-stability function. The broad proteolysis annotation is refined to the directly demonstrated protein-autoprocessing process. This does not restrict all possible FAM111A substrates to itself. The broad damage-response annotation remains refined to the two supported, non-equivalent processes.
Fine et al., PMID:23093934, Figures 3–8 and corresponding Results/Methods, establishes SV40 large T antigen association through proteomics, reciprocal co-immunoprecipitation and yeast two-hybrid domain mapping. The human FAM111A C-terminal region is tested separately from endogenous human U-2-OS protein. This interaction does not establish LT proteolysis or direct regulation of its helicase chemistry, and is retained as supported non-core generic binding under the campaign exception.
The same paper positively detects endogenous FAM111A in both nuclear and cytoplasmic U-2-OS fractions with lamin/tubulin controls. Its viral-expression experiments in human U-2-OS cells are distinct from plaque and burst assays in African green monkey CV-1P/BSC40 cells. The latter should not be described as human infection-yield assays.
Zhu et al., PMID:37607234, Figures 1, 3 and 6–7 and main Methods, independently supports human A549 viral restriction and nuclear-to-cytoplasmic redistribution during MVA infection. Tagged constructs and endogenous fractionation are separate observations. Protease-inactive constructs remain nuclear under the tested conditions, but an inactive construct forced into the cytoplasm can promote I3 loss. Thus protease-dependent export and I3 degradation are different steps. CQ and ATG16L1 perturbations support an autophagy-related route; the reported MG132 response and lack of rescue by 3-MA or bafilomycin are not evidence for a clean, exclusively resolved pathway. Direct FAM111A cleavage of I3 is not asserted. The positive cytoplasmic pools remain contextual non-core residence, not contamination by assumption.
The serine endopeptidase and single-stranded DNA-binding activities are independently measured and cooperate in the same replication-associated role. Catalytic activity supports fork processing, covalent protein–DNA repair and experimentally demonstrated autoprocessing. These three processes are not ancestor/descendant duplicates. The DNA-recognition unit records the separately measured central-domain activity and its connection to fork progression, without asserting a second pathway or a clamp-loader function.
Current GO places chromatin under cellular anatomical entity rather than protein-containing complex. Nucleus and chromatin describe the observed functional locations; no named replication complex or constitutive membership is invented. The proteolysis review is refined to autoprocessing. DNA replication remains a separately supported source-level annotation for the 2014 bulk EdU and S-phase experiments, which are not exhausted by the later stalled-fork assays; its broad ancestor is not added as another core process. Autocleavage is not presented as obligatory physiological maturation.
The official GO_REF:0000033 method describes curator placement of ancestral assertions using experimental descendant evidence. In the captured PANTHER version 19 PTHR14389 tree, the human FAM111A/Q96PZ2 leaf PTN002494620 descends from PTN002683271. Its MSA row exactly matches the 611-residue displayed protein. All five imported IBD term/donor assertions match; no relevant NOT appears in the saved IBD records or explicit tree fields on that path. This is a bounded version-specific check, not a claim about every future family release. Target self-WITH is valid descendant evidence, and donor count is not an evidence-quality score. All five source-node statuses remain SUPPORTS_TRANSFER; the historical access failure is retained once in the notes.
Public source locators are the PTHR14389 tree service, family MSA service, human Q96PZ2 gene-identity service, and PAINT IBD file. The tree and MSA were captured on 2026-10-25 and identify version 19; the IBD capture is dated 2026-10-24. These are live public endpoints and may return a later release. The relevant leaf and ancestral identifiers above identify what was checked.
Human FAM111B/Q6SJ93 is the peptidase donor, with current experimental IDA traced to PMID:32776417. Its experimental nuclear annotations were traced to PMID:36610347 and PMID:37342232, but those two donor originals were not read or cited as target experimental evidence. Direct human FAM111A localization independently supports nucleus.
The actual UniProt location records map SL-0191 to nucleus, SL-0468 to chromosome and SL-0086 to cytoplasm. The target's curated SUBCELLULAR LOCATION entry cites the actual human papers. Chromosome-to-chromatin is a specificity refinement supported by nascent-chromatin capture; chromatin is part_of chromosome, not an is_a chromosome subtype. Broad nucleus remains useful because replication foci do not exhaust the observed nuclear pools. Cytoplasm is broader than cytosol and is retained without silently changing the imported concept.
PMID:33369867(https://pmc.ncbi.nlm.nih.gov/articles/PMC7857424/) independently reports endogenous FAM111A at PCNA-positive replication sites and PCNA/PML-associated U2OS foci. The selected human-cell, construct and immunofluorescence Methods and corresponding Results were read. Nucleolar and other nuclear pools are not all recast as chromatin.
All six cited publication identities are verified. The immutable normal caches for PMID:24561620 and PMID:32165630 are abstract-only, so their unavailable flags remain true even though complete official external main texts were recovered and selectively read. The other four cited caches contain full main text and retain false flags. Five shared papers preserve both complete direct and database assessments with explicit labels; availability does not imply whole-paper reading.
The direct lane actually viewed PMID:32776417 Figure 1 and EV1 pixels, including human FAM111A WT/D439N and AEBSF controls. The database lane independently viewed the A/B controls in Figure 1G/H and EV1C–G. Other figure references identify read text/captions, not viewed pixels. No raw gel reanalysis, source-data table, supplemental sequence assembly, Text S1 or SI Appendix reading is claimed. PMID:24561620's author-manuscript heading differs from the published normal title but matches the DOI/PMCID.
The 611-residue displayed Q96PZ2 protein and independently recovered locus models remain distinct from tagged historical constructs, engineered mutants, truncations and cleavage products. Matching protein sequences do not establish identical expression or regulation across transcripts. There is no UniProt alternative_products field and none is manufactured. Supported generic interactions remain non-core under the explicit campaign exception; no new publication quotations or NEW annotations are proposed.