CDC45 manual research and annotation review

CDC45 is the structural replication factor in the human CDC45–MCM2–7–GINS (CMG) helicase. The MCM ring supplies ATP-driven translocation/unwinding; CDC45 and GINS organize its active architecture. The core therefore combines CDC45's experimentally supported ssDNA binding with a contributes_to DNA-helicase activity and CMG membership. This does not assign independent helicase, ATPase or nuclease catalysis to CDC45. All 43 source assertions, their qualifiers and partners, the three alternative products and the original 25 reference identities/titles are preserved. Six normally fetched primary sources provide corroboration; no NEW annotation is proposed.

Primary evidence and scope

PMID:22474384, human CMG biochemistry: the normal publication cache is abstract-only, but authentic external full HTML was inspected for Results/Figures 1–3 and 6 and Methods. Eleven human subunits were expressed in Sf9 cells and purified as CMG. ATPase/helicase assays and rolling-circle replication with human polymerase epsilon test the complex. HeLa chromatin co-immunoprecipitation and active preparations from 293 cells corroborate human complex membership. These are neither CDC45-alone catalysis nor yeast proteins mislabeled as human.
PMID:24293646, purified human CDC45 DNA binding: selected full Results/Figures 4–7 and expression/EMSA methods establish binding by full-length, C-terminal His6-tagged human CDC45 expressed in bacteria and assayed purified in vitro. A brief provenance anchor is “hCdc45 displayed a preference for 3' protruding strands”. The study also tests ssDNA length dependence and branched substrates. Its passive, ATP-independent strand destabilization is distinct from GO:0003678, whose definition requires ATP-driven unwinding. The tested substrates do not independently establish origin-sequence specificity.
PMID:27189187, CDC45 structure: selected full structural Results and Methods distinguish the human crystal construct (internal deletion 154–164), docking into fly CMG, and Xenopus interface-mutant replication/chromatin experiments in egg extracts. Conserved MCM/GINS interfaces support an architectural role. Loss of the conserved RecJ catalytic motifs prevents transferring nuclease activity from the ancestral fold.
PMID:32453425, human CMG/CMGA structures: selected full Results and HEK293-F expression/purification, fork-DNA and cryoEM methods show human CMG bound to DNA/ATPγS and the AND-1 interface. DNA and nucleotide contacts in MCM2–7 distinguish the motor from CDC45/GINS architecture. Structural recruitment of AND-1 does not make CDC45 a polymerase or a general signaling adaptor.
PMID:35585232, human replisome reconstitution: selected full minimal-replisome Results/Figure 1 and replication-reaction Methods show CMG-dependent leading-strand DNA synthesis. This establishes direct participation of the replication assembly, rather than interpreting a knockout phenotype as the biochemical mechanism.
PMID:27374770, Mendelian disease: the authentic abstract describes biallelic splice/substitution variants and reduced full-length transcript/protein in patient cells. This supports hypomorphic Meier-Gorlin syndrome/craniosynostosis. Full experimental Methods were not recovered; no exact mutant rescue, numerical activity deficit or alternative-product equivalence is asserted.

Decisions outside the core

The inherited PAINT annotations remain assertions of ancestral-node placement, not pairwise similarity votes. The node is PANTHER:PTN000053416; the complete historical tree/MSA/IBD reconstruction was not independently inspected. No human-specific loss of the conserved replication apparatus was established. Replication initiation, preinitiation-complex membership and structural participation in assembly are retained as core. Chromatin and origin binding are retained as non-core context, with the distinction between origin recruitment and independent sequence-selective recognition made explicit. The self-donor O75419 on the complex annotation is legitimate descendant evidence, not circularity.

The existing BIR inference is retained as a non-core repair context of the conserved replication machinery. PMID:20516198(https://pubmed.ncbi.nlm.nih.gov/20516198/) provides experimental homolog-pathway evidence for Cdc45/MCM/GINS involvement; its inspected abstract is not a direct human CDC45 BIR assay. This does not create an additional process annotation or promote repair to CDC45's principal function.

The exact Human Protein Atlas CDC45 subcellular page, including per-antibody/cell-line calls, supports nucleoplasm and cytosol. HPA000614 and CAB009459 show nucleoplasmic staining in A-431, U-251MG and U2OS; CAB009459 additionally shows cytosol in A-431/U2OS. Human nuclear-fraction evidence in PMID:21383955(https://pubmed.ncbi.nlm.nih.gov/21383955/) and CMG chromatin recovery supply replication context. The cytosolic pool has no separately established core function. Centrosome staining does not by itself resolve the ciliary basal-body annotation. The five Reactome summaries are retained as pathway context; neither claspin phosphorylation nor E2F1-mediated CDC45 transcription is assigned as a CDC45 protein activity.

Three original annotations remain UNDECIDED:

Checkpoint signaling, TAS PMID:9660782. The accessible abstract establishes ORC2 association and nuclear fractionation, but the full paper was not recovered. CLSPN association/UV response in PMID:23910567(https://pubmed.ncbi.nlm.nih.gov/23910567/) does not alone distinguish direct checkpoint signal transduction from replication machinery responding to stress. No conclusion that the original source is wrong follows.
POLA1/P09884 binding, IPI PMID:10518787. The authentic abstract explicitly reports MCM7 and polymerase-alpha p70 binding. Without the full paper or an exact target record, whether POLA1 was additionally tested remains unresolved. Neither the source partner nor its IPI evidence is rewritten.
Ciliary basal body, IDA PMID:29257953. The full cached text describes the evolutionary ciliome study and mammalian GFP-localization screening, including Table S5/Figure S1. The exact CDC45 construct/result was not recovered. A target absent from accessible main text is not evidence of an incorrect curator annotation. No wrong-gene, wrong-species or false-localization conclusion is made.

Target-specific interaction evidence

The standing ClinGen project binding instruction at the inspected main commit retains supported generic protein binding as non-core when a more informative activity is not established. It does not verify an uninspected edge. The following assessments use exact O75419/partner/PMID matches in IntAct, retaining mutation and experimental-context features. The accompanying bounded machine-field extract records source rows, exact accessions, PMID joins, assay hosts, negative flags, mutation features, query/record URLs and response/line hashes; it contains no source-comment prose. They do not claim a complete independent reanalysis of the authors' networks.

Source	Exact target evidence inspected	Interpretation
15232106	P38936; EBI-375344, in vitro protein array	Supported CDKN1A association; no inhibitor function transfer
17887956	Q92547; abstract and target pull-down/co-IP records	Human TOPBP1 association, including HeLa-S3 context
25241761	P38936; EBI-6589845, HeLa proximity ligation	Endogenous proximity; not purified binary affinity
25416956	Q53EZ4; EBI-10188186/10345943/10506346	Human-protein two-hybrid screen/validation in yeast
31515488	Q53EZ4; EBI-22145270, TTLL5 variant features with no interaction effect	Partner variants are not CDC45 substitutions
32296183	P61018, Q53EZ4, Q96D03, Q9UGI0; three exact records per pair including EBI-24686581/24437429/24272955/24320690	HuRI two-hybrid positives in yeast; no partner-function transfer
33961781	Q9BRT9 in HEK293T/HCT116; Q9HAW4 in HEK293T; EBI-54669885/54923773	CDC45-bait, spoke-expanded affinity-purification associations
39009827	O00629; EBI-54804533/55021140 plus primary Table 1 and motif-assay text	Wild-type/R157C CDC45 152–166 peptides bind engineered KPNA4 pockets; full-length KPNA7 example kept separate

The KPNA4 evidence establishes CDC45 as cargo, not a nuclear-import receptor. IntAct mutation-effect records were not treated as wild-type-only experiments; the primary peptide comparison resolves the positive interaction. For MCM2/GINS4 annotations from the abstract-only PMID:19805216(https://pubmed.ncbi.nlm.nih.gov/19805216/), exact source constructs remain uninspected, but human CMG structure/reconstitution independently corroborates physical-complex association. RPA association in the PMID:23910567 abstract does not establish a purified RPA2-only assay. Selected full Figure 4/Results in PMID:26167883(https://pubmed.ncbi.nlm.nih.gov/26167883/) establish CDC45–MCM2 co-recovery without transferring MCM2 histone-chaperone activity.

Research provenance

Normal gene/publication retrieval seeded the review and the original source bytes are preserved. Existing main publication/Reactome caches were selected by exact Git-blob identity; fresh normal variants remain separate. A genuine Falcon attempt with an explicit perplexity-lite fallback timed out during client/dependency setup at 90 seconds per attempt and produced no research report. These manual notes are not a provider-generated document. Sources marked abstract-only remain so even where an external full-text access was separately successful. Failed/challenge responses are not treated as article content. The notes and YAML use brief, nonduplicated provenance quotations with aggregate per-source counts checked across both authored files.