FASTKD2

Summary: FASTKD2 participates in mitochondrial RNA maturation as an RNA-binding regulatory factor.

Reason: The actual PTHR21228 v19 tree places human Q9NYY8 (PTN002501556) below IBD node PTN000477113; its aligned sequence matches the 710-residue product, and no family NOT assertion interrupts this path. The GO:0000963 IBD is dated 20260529. Human iCLIP and loss-of-function analysis show binding of mitochondrial transcripts and altered mature ND6/16S RNA levels and processing (PMID:26370583). This is participation in RNA maturation, not a claim of autonomous nuclease activity. The inference is judged at the actual ancestral node, not by donor count; the target evidence supports retention.

Summary: FASTKD2 is an experimentally supported RNA-binding protein.

Reason: The human mRNA-interactome study identifies FASTKD2 and individually validates RNA association by UV crosslinking, stringent immunopurification of tagged protein and oligo(dT) detection (Figure 3C and associated Methods). This is positive RNA-binding evidence, although the N-terminal fusion does not establish native mitochondrial targeting. Later mitochondrial iCLIP detects rRNA and mRNA contacts, so the broader RNA-binding term retains substrate coverage beyond rRNA binding. No kinase activity is inferred from the protein name.

Summary: FASTKD2 occurs in the human mRNA-bound proteome.

Reason: The author's original-study thesis Table S1 lists FASTKD2 as a class-I mRNA-capture protein (PDF page 142); Methods describe UV-dependent oligo(dT) capture under denaturing conditions in HEK293 cells with isotope-label controls. This supports the imported high-throughput RNA-binding assertion. An individual FASTKD2 PAR-CLIP experiment is not established here. The broad term remains useful alongside independently demonstrated rRNA and mitochondrial mRNA contacts.

Summary: FASTKD2 is an RNA-binding protein with both ribosomal and messenger RNA targets.

Reason: The actual PTHR21228 v19 tree places human Q9NYY8 (PTN002501556) below IBD node PTN000477113; its aligned sequence matches the 710-residue product, and no family NOT assertion interrupts this path. The RNA-binding IBD is dated 20250903 and legitimately includes Q9NYY8 among the descendant experimental sources. Target self-WITH is not circular. Human iCLIP detects rRNA and ND6 mRNA contacts, so this broader molecular function remains useful beyond the separately annotated rRNA binding rather than merely duplicating a single substrate class (PMID:26370583). No kinase activity is inferred from the family name.

Summary: A screened FASTKD2–LAMP2 interaction is retained as non-core.

Reason: The original human-protein yeast two-hybrid network contains the FASTKD2–LAMP2 edge (Table S2, NN1.0 row 3476); current IntAct records resolve the partner as P13473-2 and record pooling, array and validated two-hybrid assays. The gene-level author table does not independently resolve that partner suffix. This is a positive screen interaction, without establishing a lysosomal, autophagic or membrane-trafficking function for FASTKD2. See FASTKD2-binding-records.yaml for the exact joins.

Summary: A screened FASTKD2–SH3GLB1 interaction is retained as non-core.

Reason: The original human-protein yeast two-hybrid network contains the FASTKD2–SH3GLB1 edge (Table S2, NN1.0 row 9767), joined to Q9Y371 in the current IntAct pooling, array and validated two-hybrid records. The study does not establish a physiological endophilin or membrane-remodeling activity for FASTKD2. See FASTKD2-binding-records.yaml for the exact joins.

Summary: The MaMTH-HTS FASTKD2–CFTR screen hit is retained as non-core.

Reason: The author dataset identifies FASTKD2 with wild-type CFTR: Dataset EV2 wt-CFTR row 96 has the prey in one of two technical sequencing replicates, and Dataset EV4 row 85 retains the Q9NYY8–P13569 interaction. The assay uses tagged proteins and a split-ubiquitin reporter in HEK293 cells; FASTKD2 has an N-terminal prey tag. This does not establish native submitochondrial access, an individually validated channel-regulatory effect, or CFTR function for FASTKD2. See FASTKD2-binding-records.yaml.

Summary: CFTR-proximal detection of FASTKD2 is retained with assay limits.

Reason: The original Supplementary Table S1 identifies Q9NYY8 in the CFTR proximity-labeling datasets. TurboID enrichment versus control is positive (p=0.000637; log2 fold change 3.16679), while its SAINT FDR is 0.0962, above the study's 1% high-confidence threshold; the APEX2 comparison is not enriched. IntAct records the CFTR–FASTKD2 pair as proximity. Retention follows the project exception for supported generic interaction records, with no purified binary binding, stable physiological complex or CFTR-regulatory activity asserted. See FASTKD2-binding-records.yaml.

Summary: Human proteomics supports mitochondrial localization of FASTKD2.

Reason: FASTKD2/Q9NYY8 is included in MitoCoP Table S1B row 337 and has concordant positive evidence in subtractive proteomics (Table S2B row 6716, class 1), spatial proteomics (S2E row 7627, cluster 1), and TOMM40 importomics (S3A row 5777, class 1). This is direct organelle-level evidence. The protein group includes alternative accessions and does not establish an isoform-specific or matrix-specific localization. Mitochondrion remains an informative supported organelle annotation.

Summary: Antibody imaging supports mitochondrial localization of human FASTKD2.

Reason: The actual HPA ENSG00000118246 record uses HPA067191 and reports supported mitochondrion localization in CACO-2, Hep-G2, U2OS and KOLF2.1J assays. The selected U2OS image and assay metadata were inspected. These observations support the organelle-level term without resolving matrix, RNA granules or a natural isoform. Mitochondrion remains informative under the ORGANELLES policy despite finer compartment annotations.

Summary: FASTKD2 is imported into mitochondria.

Reason: The original Figure 4 and associated experiments show tagged human FASTKD2 colocalization with mitochondrial mtSSB in COS7 cells and membrane-potential-dependent import into isolated HeLa mitochondria, with protection from added proteinase K until membrane disruption. These results support mitochondrial localization. The proposed cleavage coordinate and inner-membrane topology were predictions, not experimentally mapped processing or topology; no product boundary is inferred.

Summary: Human FASTKD2 localizes to mitochondria in the FASTK-family study.

Reason: Although FASTKD3 is the paper's main functional focus, the actual Results assay human FASTKD1–5 C-terminal YFP fusions, including FASTKD2, and report mitochondrial-antigen overlap in COS7 cells, with comparable localization in stable U2OS cells (Figure S4). The direct FASTKD2 localization evidence is retained; FASTKD3-specific respiratory phenotypes and interaction functions are not transferred.

Summary: The ARBA mitochondrial prediction agrees with independent human localization evidence.

Reason: ARBA00026962 has GO:0005739 as its actual consequent. The current 380 condition sets were inspected, but complete eligibility of Q9NYY8 under a particular clause was not independently reproduced: the captured target InterPro/PANTHER entries do not themselves identify a matching clause and FunFam conditions remain unresolved. This source limit does not overturn the positive human mitochondrial localization established by HPA and the human FASTKD2-GFP experiment in PMID:26370583. Retain the biologically supported organelle claim without treating an electronic prediction as a new assay.

Summary: The inherited matrix placement is consistent with FASTKD2 mitochondrial RNA biology.

Reason: The actual PTHR21228 v19 tree places human Q9NYY8 (PTN002501556) below IBD node PTN000477113; its aligned sequence matches the 710-residue product, and no family NOT assertion interrupts this path. The matrix IBD is dated 20251219 and includes direct target evidence Q9NYY8 among its descendant sources; that self-WITH is expected. Human FASTKD2-associated mitochondrial RNA granules and the RNA-processing module provide a coherent matrix context (PMID:25683715; PMID:27974379). Matrix remains an informative location for the broader RNA-handling compartment rather than restricting the protein to the fraction that colocalizes with nucleoids. This does not establish a universal isoform-specific location or a cleavage boundary.

Summary: FASTKD2 functions in the mitochondrial matrix, including mitochondrial RNA granules.

Reason: The original FASTKD2 mitochondrial immunopurification/BioID and RNA-granule localization evidence is consistent with the matrix compartment (PMID:27974379, Results/Figure 1, expanded Figure EV1 and localization Methods). Independent endogenous FASTKD2 localization in PMID:25683715 and the curated UniProt CC SUBCELLULAR LOCATION statement support this placement. Accept the experimental curator's matrix assignment consistently with the other positive matrix annotations. These sources do not resolve freely soluble matrix residence from association with the matrix-facing inner membrane, and RNA-granule localization is partial rather than exclusive. That subcompartment/topology limit does not negate the broader matrix claim; no membrane topology or cleavage coordinate is inferred.

Summary: The combined orthology and curated-location mappings support mitochondrial matrix placement.

Reason: GO_REF:0000120 merges electronic methods. The SL-0170 mapping is exactly GO:0005759, and Ensembl identifies ENSMUSP00000027103/Q922E6 as the mouse one-to-one FASTKD2 ortholog. Its current GOA includes a genuine PMID:29617655 matrix IDA source, separately from propagated human-derived rows; this is not an exclusively circular chain. The inspected mouse paper shows Fastkd2 in heart mitochondrial fractions and with the PTCD1/RPUSD4 module, while human RNA-granule/module evidence supports the compartment (PMID:25683715; PMID:27974379). These are related localization inferences, not independent measurements of human isoform cleavage.

Summary: Reactome explicitly locates the FASTKD2 protein component in mitochondrial matrix.

Reason: Actual recursive Reactome objects establish the route 9837212 input 9837146. Entity 9837146 is linked through referenceEntity 229275 to UniProt Q9NYY8 and has compartment GO:0005759. The event also produces complex 9837175 containing the same protein and an RNA set. This is a valid cellular-component source even when the event title foregrounds another enzyme. The model represents residues 1–710, which is not experimental evidence of a mature cleavage boundary.

Summary: Reactome explicitly locates the FASTKD2 protein component in mitochondrial matrix.

Reason: Actual recursive Reactome objects establish the route 9936243 catalystActivity 9936263 physicalEntity 9936292 hasComponent 9837146. Entity 9837146 is linked through referenceEntity 229275 to UniProt Q9NYY8 and has compartment GO:0005759. The catalyst entity is the pseudouridylation module; the named chemistry is attributed to RPUSD4, not to an autonomous FASTKD2 enzyme. This is a valid cellular-component source even when the event title foregrounds another enzyme. The model represents residues 1–710, which is not experimental evidence of a mature cleavage boundary.

Summary: Reactome explicitly locates the FASTKD2 protein component in mitochondrial matrix.

Reason: Actual recursive Reactome objects establish the route 9937369 catalystActivity 9937392 physicalEntity 9936292 hasComponent 9837146. Entity 9837146 is linked through referenceEntity 229275 to UniProt Q9NYY8 and has compartment GO:0005759. The named mRNA chemistry belongs to the RPUSD3/TRUB2 module context and does not establish a FASTKD2 pseudouridine synthase activity. This is a valid cellular-component source even when the event title foregrounds another enzyme. The model represents residues 1–710, which is not experimental evidence of a mature cleavage boundary.

Summary: RNA processing is supported but the mitochondrial term captures the relevant process more precisely.

Reason: The original study combines mitochondrial RNA contacts with Northern-blot changes in ND6 processing and steady-state levels, selective 16S-rRNA loss and altered 7S RNA after FASTKD2 depletion. The retained mitochondrial RNA processing assertion (GO:0000963) is a verified strict descendant of RNA processing (GO:0006396) and covers this biological scope. The work does not establish an intrinsic nuclease reaction, a target half-life, or a non-mitochondrial RNA-processing function.

Summary: The preliminary apoptosis phenotypes do not establish FASTKD2's participation in the apoptotic process.

Reason: PMID:18771761 Figure 5 and its Results show reduced staurosporine-induced nuclear fragmentation in patient fibroblasts and fragmentation after FASTKD2 overexpression. The authors explicitly leave specific apoptotic action versus nonspecific overexpression toxicity unresolved. Crucially, respiratory-chain complex activities were reported as normal in the tested mutant fibroblasts; the COX deficiency was observed in skeletal muscle. The phenotype therefore cannot be declared a demonstrated consequence of OXPHOS failure in those fibroblasts. Nonetheless, these preliminary response and toxicity observations do not identify a normal FASTKD2 contribution to apoptotic signaling or execution. The process assignment overshoots the resolved mechanism, while the positive experimental phenotype is retained and a possible regulatory role is not excluded. UniProt CC FUNCTION likewise describes the apoptosis role tentatively.

Summary: FASTKD2 binds mitochondrial ribosomal RNA.

Reason: In human 143B cells, UV crosslinking, FASTKD2 immunoprecipitation and RNA sequencing enriched the 16S mitochondrial rRNA relative to naked-bead controls (Figure 5 and associated Methods). This supports rRNA binding without claiming a measured equilibrium affinity or kinase/nuclease activity. Independent mitochondrial iCLIP is consistent with this substrate class.

Summary: FASTKD2 supports mitochondrial translation, with RNA handling and ribosome biogenesis as its resolved core roles.

Reason: PMID:26370583 Figure 4D and the pulse-labeling Methods directly measure reduced mitochondrial protein synthesis after FASTKD2 RNAi or CRISPR depletion; emetine excludes cytoplasmic translation. Crucially, RNAi reduced translation without appreciably changing RNR2 (16S rRNA) steady-state abundance or processing on the measured Northern blots. The translation phenotype therefore cannot be attributed solely to observed 16S-rRNA depletion. The study also establishes RNA contacts and transcript defects, while PMID:25683715 connects FASTKD2 to 16S rRNA and large-subunit assembly. Retain the positive translation role as non-core to prioritize those resolved RNA-handling and assembly activities in the synthesis. The proximal translation mechanism remains unresolved: these assays do not isolate a translation step, but that limitation does not establish absence of direct participation. This is a core-prioritization judgment, not denial of UniProt's translation requirement or a claim that all effects are secondary to rRNA loss. Noncatalytic translation contributions remain possible.

Summary: FASTKD2 is concentrated in mitochondrial RNA granules.

Reason: The actual PTHR21228 v19 tree places human Q9NYY8 (PTN002501556) below IBD node PTN000477113; its aligned sequence matches the 710-residue product, and no family NOT assertion interrupts this path. The granule IBD is dated 20250903 and includes legitimate target experimental evidence Q9NYY8. In PMID:25683715, endogenous-protein immunofluorescence in 143B cells shows substantial overlap with GRSF1 and newly synthesized BrU-labeled mitochondrial RNA. This supports the granule location while preserving the reported partial nucleoid overlap and diffraction-limited resolution. It is not a claim that every FASTKD2 molecule is permanently in one granule.

Summary: FASTKD2 is a mitochondrial ribonucleoprotein-granule component.

Reason: Endogenous FASTKD2 immunofluorescence in human 143B cells overlaps GRSF1 and newly synthesized BrU-labeled mitochondrial RNA (Figure 1). The study places approximately 50–90% of the tested proteins' foci in these RNA-granule markers. Localization is partial and dynamic; the annotation does not imply exclusive granule residence. The broader ribonucleoprotein-granule term accurately describes the mitochondrial RNA granules assayed.

Summary: Partial overlap with mitochondrial nucleoids is retained as a secondary localization.

Reason: The original Figure 1 and Results report that approximately 30–40% of FASTKD2/DDX28/DHX30 foci colocalize with mtDNA nucleoids, while most RNA-granule foci are distinct. This is positive partial localization evidence, not a negative nucleoid result. Diffraction-limited overlap cannot resolve physical proximity from stable nucleoid incorporation, and does not establish DNA binding or a DNA-maintenance function. RNA-granule residence is the more prominent experimentally supported localization.

Summary: Partial nucleoid association is retained as a secondary localization.

Reason: The SL-0269 mapping is exactly GO:0042645, and the official human UniProt location statement links to PMID:20869947 and PMID:25683715. The operative PMID:25683715 endogenous 143B-cell imaging reports partial anti-DNA nucleoid colocalization for the assayed RNA-granule proteins, while the majority of granules are distinguishable from nucleoids. Retain the positive association with those spatial-resolution limits; granule predominance is not evidence of absence from the nucleoid region. This does not imply DNA binding, a nucleoid structural role, or exclusive localization. The mapped and direct annotations describe the same partial localization; both are kept as non-core because RNA-granule residence is the more prominent observed localization.

Summary: Conserved regulation of mitochondrial mRNA fate is supported by the ancestral placement and target ND6 evidence.

Reason: The actual PTHR21228 v19 tree places human Q9NYY8 (PTN002501556) below IBD node PTN000477113; its aligned sequence matches the 710-residue product, and no family NOT assertion interrupts this path. The GO:0044528 IBD is dated 20210319 and is experimentally grounded in FASTKD3 (PMID:27789713) and FASTKD1/TBRG4 (PMID:28335001). FASTKD3 knockout/rescue and transcription-block experiments directly test mRNA decay; FASTKD1/4 have distinct and sometimes opposing transcript effects. The term includes both stabilization and destabilization. Human FASTKD2 binds mature ND6 coding-region RNA and its loss depletes mature ND6 (PMID:26370583); these target data corroborate the inherited regulatory claim but do not by themselves measure an isolated FASTKD2-dependent half-life. An indirect ND6 decrease secondary to impaired large-subunit assembly or translation remains possible, as discussed in PMID:29617655; target contact and abundance are not an isolated decay assay. Predominant 16S-rRNA/assembly biology does not demonstrate loss of the ancestral mRNA function.

Summary: The actual FAST_1 domain mapping is consistent with mitochondrial mRNA regulation.

Reason: The current IPR010622 record maps to GO:0044528, and the target match spans residues 456–529 of Q9NYY8. Domain membership alone does not specify a particular RNA or catalytic chemistry. The mapping is biologically consistent with the actual PAINT inheritance and human FASTKD2 ND6-binding/loss phenotype, with donor turnover evidence from PMID:27789713 and PMID:28335001. Retain the broad regulatory process while distinguishing inferred stability regulation from a target-specific measured decay constant; no kinase activity is inferred from the historic domain name.

Summary: Orthology transfers a positive modification-module association, but FASTKD2-specific participation in the chemical step remains unresolved.

Reason: The Ensembl one-to-one mouse donor Q922E6/ENSMUSP00000027103 has a real FlyBase IPI GO:1900864 annotation to PMID:29617655 with PTCD1/Q8C2E4. The full paper reports mouse-heart PTCD1–Fastkd2 co-immunoprecipitation, co-fractionation with RPUSD4, and loss of 16S-rRNA pseudouridylation after PTCD1 disruption. The perturbation directly testing chemical modification is PTCD1 loss, not Fastkd2 loss. Human studies place FASTKD2 in an RNA-handling module with pseudouridine synthases (PMID:27667664; PMID:27974379), but the inspected experiments do not isolate whether FASTKD2 supplies a structural or regulatory contribution specifically to modification. Such a noncatalytic contribution remains possible; lack of a catalytic domain is not a disproof. Retain source-specific uncertainty rather than equating general RNA maturation or complex association with a demonstrated chemical-modification role, and do not infer a wrong-gene annotation or biological absence. UniProt CC SUBUNIT's regulatory protein-RNA complex statement is supported by the positive NGRN/WBSCR16 immunoprecipitation and RNase-resistant interactions in PMID:27667664 (Results/Figure 6; Discussion leaves single versus overlapping complexes unresolved). PMID:27974379 Figure 3D instead finds largely monomeric migration under its 2D-gel conditions and proposes transient associations. These observations support physical/module association without establishing one stable obligate holocomplex, and neither isolates a FASTKD2-specific contribution to modification. In PMID:29617655, RPUSD4 is the synthase and PTCD1 loss is the tested perturbation; a noncatalytic FASTKD2 contribution remains plausible but unseparated. Keep the uncertainty rather than asserting that association is sufficient or that FASTKD2 cannot contribute.

Summary: FASTKD2 contributes to mitochondrial large-ribosomal-subunit assembly.

Reason: FASTKD2 depletion reduces 16S rRNA, assembled large subunits and monosomes and redistributes unassembled ribosomal proteins on sucrose gradients (Figures 3–4); the same study directly detects FASTKD2 binding to 16S rRNA (Figure 5). This supports an RNA-associated assembly role. FASTKD2 itself largely remains in light fractions, so the evidence does not make it a stable structural constituent of the mature mitoribosome.