Reviewed the seeded human Q96PY6 annotation records, fetched UniProt description and features, original cached publications, and additional primary studies. No custom sequence analysis was necessary: the catalytic domain and direct biochemical activity are established. Gene-specific summaries were checked against primary evidence rather than treated as sources of experimental claims.
NEK1 is a serine/threonine protein kinase with distinct centriolar/ciliary and genome-maintenance roles. The most informative recent advance is kinase-dependent control of centriole length. In human cells, NEK1 loss or acute degradation causes centriolar microtubule hyperextension despite continued CP110-CEP97 occupancy; CP110 codepletion enhances the phenotype. Wild-type rescue and the failure of kinase-dead D146A rescue support an enzymatic role. PMID:42265085 This supports NEW GO:1903723 negative regulation of centriole elongation, verified against OLS and QuickGO, rather than a claim that NEK1 is part of the CP110-CEP97 complex.
Centriole association is dynamic. The 2026 localization study places NEK1 at distal centriole ends in cycling cells, but reports CEP78-dependent removal from the basal body during ciliogenesis. Earlier centrosomal/pericentriolar and basal-body observations should be interpreted with cell-cycle and ciliogenesis stage in mind, not rejected as contradictory. PMID:42265085 Added NEW GO:0005814 centriole with this explicit timing caveat.
The requirement for NEK1 in cilium assembly remains well supported. Patient material shows reduced cilium number and altered morphology, and knockout/rescue work establishes that kinase activity is required. PMID:21211617 PMID:37188479. The exact phosphorylation targets mediating length control and ciliogenesis are still unresolved, including in the 2026 paper.
CFAP410, called C21ORF2 in the literature, forms a stable cellular complex with NEK1. Older work recovered SPATA7 with these proteins and showed basal-body colocalization, but later endogenous-complex experiments found CFAP410 to be the dominant interactor under the tested unchallenged conditions. The evidence supports a functionally important NEK1-CFAP410 association; it does not establish an obligate trimer or a purified direct binding constant. PMID:26167768 PMID:37188479
NEK1 kinase activity also supports DNA-damage checkpoint signaling. Human-cell depletion and rescue studies support regulation of ATR-ATRIP stability and signaling competence. Do not depict ATR Thr1989 as a direct NEK1 phosphorylation site: the relevant study explicitly tested and excluded that mechanism. PMID:23345434 The older damage-induced activation study and the 2013 priming study use different experimental settings; a uniform model of damage-induced kinase activation is not assumed.
NEK1 promotes homologous-recombination repair independently of whether one particular substrate assignment is resolved. Human U2OS reporter experiments directly show the repair defect after NEK1 depletion, supporting NEW GO:0000724 double-strand break repair via homologous recombination. PMID:26290490. The 2016 report attributed this role to RAD54 Ser572 phosphorylation. A 2023 primary study explicitly notes a challenge raised in a 2022 preprint; the review records this at the finding level as DISPUTED, not as a definitive refutation. PMID:27264870 PMID:37188479 PMID:37439356 was inspected by the coordinating reviewer and does not establish a peer-reviewed refutation of Ser572 phosphorylation; it is not used for that claim.
A mitochondrial-associated pool of NEK1 phosphorylates VDAC1 Ser193 and limits injury-induced mitochondrial apoptotic signaling. Direct substrate assays and mutant phenotypes support the activity, and fractionation/microscopy support NEW GO:0005739 mitochondrion. Perturbation and caspase activation support NEW GO:0043066 negative regulation of apoptotic process. PMID:19158487 PMID:20230784 The core description avoids the older model's stronger assertion that VDAC1 is an obligatory structural component of the permeability-transition pore.
The Reactome cytosolic ME1 reaction provides an additional compartment-specific substrate example; its primary source is PMID:31735643. It is retained as support for the cytosol annotation rather than used to elevate broad tumor-metabolism phenotypes into additional core functions. [Reactome:R-HSA-9861642 "phosphorylates cytosolic malic enzyme 1 (ME1) at Ser-336, inhibiting its catalytic activity"]
Recent patient-derived and iPSC motor-neuron work corroborates ciliary abnormalities and altered signaling in NEK1-associated ALS. Its observed calcium/AURKA/HDAC6 pathway changes do not establish that each affected downstream protein is a direct NEK1 substrate. PMID:40389989 Disease susceptibility is biological context, not a molecular-function annotation.
All 37 seeded records were reviewed individually and their original GO identifiers, evidence codes, references, qualifiers and interaction partners preserved. Broad but valid protein kinase and ATP-binding annotations were retained. Generic kinase activity from the VDAC1 assay was refined to protein serine kinase activity.
Five generic protein-binding rows whose GOA partner is UniProtKB:Q04917 were refined to the existing, more informative GO:0071889 14-3-3 protein binding. The fetched UniProt record explicitly identifies Q04917 as YWHAH. This refinement uses the original curated partner identity and does not infer phosphorylation of YWHAH. The pre-existing orthology-supported 14-3-3 binding annotations were retained as non-core regulatory interactions.
Other generic protein-binding rows were removed solely because they do not describe a specific molecular function. Their interactions with FEZ1, FEZ2, LRRK2, CFAP410 and ALS2 were not adjudicated as false. No adapter activity, substrate phosphorylation or pathway execution was inferred from association alone. The initially abstract-only LRRK2 screen PMID:24510904 was successfully refetched with full text before completion. Some interaction partners occur in large-scale data rather than the narrative; the GOA assertion was preserved as provenance and the removal is semantic, not a claim of experimental error.
The current HPA page was checked directly: https://www.proteinatlas.org/ENSG00000137601-NEK1/subcellular (accessed 2026-09-12). It lists nucleoplasm as main localization and cytosol and centriolar satellites as supported additional locations, supporting the three existing immunofluorescence-derived annotations. Primary-cilium staining is separately rated uncertain by HPA and was not added on the basis of this resource.
PANTHER IBA protein kinase activity is supported by demonstrated NEK1 catalysis. The presence of the target among WITH/FROM evidence is legitimate descendant support for an ancestral PAINT assertion and was not interpreted as circularity.
New author-supplied GO terms were checked through OLS search and QuickGO definitions. All supporting text in the YAML is copied from the referenced immutable publication or Reactome cache; wrapped abstracts are preserved verbatim. Publication titles were updated from fetched records where initialization had left placeholders. The review contains 42 annotations, including five NEW proposals, and three activity-oriented core functions.
The completed Falcon research report and its artifact were read after primary-source curation. The report correctly emphasizes the NEK1-CFAP410 complex and unsettled substrate spectrum, but its statement that NEK1 phosphorylates “ATRIP at T1989” is incorrect. The primary study tests ATR Thr1989, not an ATRIP Thr1989 substrate, and explicitly excludes direct NEK1 phosphorylation at that site. PMID:23345434 The generated research files remain unchanged; this correction governs the curated YAML.
Falcon also identified the 2023 motor-neuron study, which was fetched and checked directly. NEK1 knockdown, kinase inhibition and a patient variant perturb microtubule homeostasis and nuclear import in human iPSC-derived motor neurons; biochemical assays report alpha-tubulin and importin-beta1 phosphorylation in vitro. Endogenous site-specific phosphorylation and the causal route from those sites to the cellular phenotypes remain to be demonstrated. PMID:37585529 This is included as neuronal context in the description and as an unresolved substrate question, without assigning NEK1 import-receptor activity or asserting direct physiological substrate sites.
The research report predates the specifically curated June 2026 centriole-length study in its mechanistic synthesis. The primary 2026 results therefore take precedence for current centriole-localization timing and the additional length-control process. Literature residue numbering can refer to the 1286-residue experimental form, whereas the fetched UniProt canonical sequence is 1258 residues; residue labels in notes identify the source constructs, and no isoform coordinate equivalence was invented.
The NEK1 plasma-membrane appearance in the June 2026 study belongs to the ReLo interaction assay: a pleckstrin-homology domain artificially targets the bait to the plasma membrane in Drosophila S2R+ cells. PMID:42265085 The observation establishes neither native plasma-membrane residence of human NEK1 nor recruitment of the endogenous protein to that membrane. The paper also explicitly identified nonspecific CP110 binding to the PH-mCherry control; this is another reason not to infer a native NEK1-CP110 complex from the ReLo screen alone. PMID:42265085 No plasma-membrane annotation is proposed for NEK1 from this experiment.
A second membrane-assay example appears in the VDAC1 phosphorylation study: recombinant VDAC1 was inserted into liposomes and isolated erythrocyte membranes for transport measurements. PMID:20230784 This artificial insertion does not establish native plasma-membrane localization of NEK1 or VDAC1. It is distinct from the mitochondrial association measured in intact-cell microscopy and fractionation in the earlier study. PMID:19158487
The same 2026 study directly detects NEK1 along the axoneme of serum-starved RPE1 cells by ultrastructure expansion microscopy, even while basal-body localization is lost. PMID:42265085 Added NEW GO:0005930 axoneme after OLS search and QuickGO definition verification. The evidence shows different compartment occupancy; it does not track molecules moving from the mother centriole to the axoneme. Degradation versus redistribution is an explicit unresolved question in the paper.
Kinase-dead NEK1 D146A fails to fully restore centriole-length control, but its phenotype is milder than complete knockout. The interpretation is therefore kinase-dependent control with a possible additional noncatalytic contribution, not proof that all relevant activity is catalytic. The review rationale now says “fails to restore it fully.” PMID:42265085
The primary bioRxiv API was queried for both relevant preprints on 2026-09-13:
published: NA. The abstract reports failed replication of NEK1-dependent RAD54 Ser572 phosphorylation and HR effects. Exact abstract snippet: “Specifically, we found that Nek1 does not phosphorylate RAD54-S572 as was reported.”published: NA. Exact abstract snippet: “Here, we establish that Nek1 regulates the phosphorylation of Rad54 at Ser572 (S572) to promote HR in vivo in adult mice and in vitro in fibroblasts derived from such mice.” The same abstract reports NEK3/NEK5-dependent regulation in embryonic fibroblasts instead of NEK1.These are identified as preprints, and no journal publication was established by the checked provider metadata. Full-text XML requests for both returned HTTP 429; the new audit therefore verifies their abstracts and publication metadata, not their detailed methods or underlying data. The later adult-mouse result counters treating the 2022 challenge as definitive, but does not establish the physiological human substrate assignment. DISPUTED is retained. No evolutionary transfer from the mouse developmental result or definitive NEK1-to-RAD54 catalytic edge is added. The original-author online rebuttal was encountered but is not used as decisive experimental evidence.
Fetched the independent follow-up [PMID:31914854, "The TLK1/Nek1 axis contributes to mitochondrial integrity and apoptosis prevention via phosphorylation of VDAC1"]. It reports altered VDAC1 phosphorylation/stability and mitochondrial survival in human-cell models. PMID:31914854 Its idea that recruitment of NEK1 to nuclear DNA-damage sites depletes its mitochondrial pool is introduced explicitly as a model. It should remain a testable redistribution hypothesis, not a measured molecular itinerary. Neither this study nor the older reconstitution assay warrants converting mitochondrial association into plasma-membrane localization or making NEK1 a structural component of the mitochondrial permeability-transition pore.
The ATR/ATRIP distinction is settled by the directly tested primary experiment: NEK1 helps prime ATR-ATRIP, but the relevant site is ATR Thr1989 autophosphorylation, not an ATRIP site or a directly phosphorylated NEK1 substrate. PMID:23345434
Read the complete generated NEK1-deep-research-openscientist.md after the independent audit. It corroborates the distinction between ATR priming and direct substrate phosphorylation, the need to separate in-vitro neuronal substrate assays from physiological site dependence, and the unresolved RAD54 claim, including the newly identified December 2025 mouse preprint. It is cited only as a secondary synthesis alongside directly checked primary evidence. Its declared reliance on abstracts and open-access summaries, rather than direct inspection of all relevant figures and supplements, materially limits several stronger conclusions:
The OpenScientist files and artifacts are preserved unchanged. The reference-level assessment records these limitations so future readers do not treat the provider's stronger statements as independently validated facts. After integration, the curated review contains 43 annotations (37 original records and six NEW proposals), with three core functions and no PENDING decisions.
The five 14-3-3 binding refinements now rely explicitly on their preserved GOA IPI partner identity (Q04917/YWHAH); the unrelated potential-substrate quote from PMID:19158487 has been removed. The NEW anti-apoptosis annotation now quotes the Figure 2F caspase result and its wild-type/GFP controls rather than the unshown cytochrome-c observation. The broad cilium-assembly row is refined to GO:1905515 non-motile cilium assembly, harmonizing the primary-cilium evidence with CFAP410. Suggested experiments now state their hypotheses and assay types.
The fetched GOA/UniProt records, primary-publication caches, Reactome entry and unchanged provider reports are included in the PR so reviewers and CI can reproduce the local evidence checks. The earlier exclusion followed the PR template, but left required file references unresolved. Cached publication text was generated by the fetch tooling and was not manually edited.