Multidomain peripheral protein of early endosomes, built from an N-terminal pseudokinase domain, a central Tre2-Bub2-Cdc16 (TBC) domain and a C-terminal rhodanese-like domain. Despite its name, TBCK is not a protein kinase: the kinase-fold domain lacks the glycine-rich loop and the VAIK, HRD and DFG motifs, purified full-length human protein binds neither ATP nor GTP and hydrolyses neither, placing it among the class I pseudokinases. The catalytically competent module is the TBC domain, which retains the arginine and glutamine fingers of Rab GTPase-activating proteins; TBCK is the catalytic core of a heterotrimer with PPP1R21 and FERRY3 that acts as a GTPase-activating protein for RAB5, and its loss leaves RAB5 constitutively GTP-loaded, enlarges early endosomes and blocks their conversion to late endosomes. TBCK is also one of the five subunits of the FERRY complex, a RAB5A effector that tethers mRNAs and ribosomes to early endosomes, although the mRNA contacts are made by other subunits. Reported effects of TBCK depletion on mTOR signalling, cell size, proliferation and actin organisation are not consistently reproduced across cell types and remain disputed. Biallelic loss-of-function variants cause TBCK encephaloneuronopathy (infantile hypotonia with psychomotor retardation and characteristic facies 3), a severe autosomal recessive neurodevelopmental and neurodegenerative disorder.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005096 GTPase activator activity | IBA GO_REF:0000033 | ACCEPT | Summary: The phylogenetic assertion matches what is now the best-supported molecular function of TBCK: the TBC domain is an active Rab-GAP module, and TBCK is the catalytic core of a RAB5-directed GAP complex. Reason: Core molecular function, and the better-supported of the two activities the gene's annotations assert. Three independent lines converge: the TBC domain retains the canonical GAP arginine and glutamine fingers; the recurrent disease missense allele p.Arg511His removes the arginine finger; and biochemical/cell work identifies a TBCK-PPP1R21-FERRY3 heterotrimer with specific GAP activity toward RAB5, whose loss produces exactly the RAB5-hyperactivation phenotype predicted. Note that GO merged the former 'Rab GTPase activator activity' into this term, so GO:0005096 is the most specific available MF; the Rab5 specificity cannot currently be expressed in the MF itself. Supporting Evidence: PMID:41789809 This complex functions as a specific GTPase-activating protein (GAP) for RAB5. PMID:41789809 TBCK deficiency or missense mutations of its key residues in the RABGAP-TBC domain lead to constitutive RAB5 hyperactivation, which blocks the transition from early to late endosomes and results in the formation of massively enlarged RAB5-positive endosomes. PMID:42546825 The TBC domain of TBCK retains conserved arginine and glutamine residues characteristic of GTPase-activating proteins (GAPs), suggesting that TBCK may function as a GAP toward small GTPases PMID:27040692 Structural analysis implicated Arg511 as a required residue for Rab-GAP function |
| GO:0005524 ATP binding | IEA GO_REF:0000002 | REMOVE | Summary: Signature-based transfer from the protein-kinase-domain InterPro entry (IPR000719). Purified full-length human TBCK does not bind ATP, ADP, GTP or GDP. Reason: An electronic inference from a kinase-fold signature that direct biochemistry contradicts. Differential scanning fluorimetry over ATP, GTP, ADP, GDP and Mg2+ produced no ligand-induced thermal shift outside experimental error, and the domain lacks the glycine-rich loop and VAIK motif that would form the nucleotide pocket. UniProt itself annotates the kinase domain as predicted catalytically inactive. Removing an IEA signature mapping that is demonstrably wrong is the intended use of this action; no experimental annotation is being overruled here. Supporting Evidence: PMID:42546825 Biochemical and biophysical analyses reveal that the catalytically inactive pseudokinase domain of TBCK lacks nucleotide binding, consistent with the absence of the canonical VAIK, HRD, and DFG motifs required for catalysis. PMID:42546825 These results, therefore, indicate that none of the tested nucleotides measurably stabilize or destabilize the global thermal stability of TBCK under the conditions examined, consistent with TBCK functioning as a class I pseudokinase. |
| GO:0005737 cytoplasm | IEA GO_REF:0000044 | ACCEPT | Summary: TBCK is a soluble, peripherally membrane-associated protein of the cytoplasm. Reason: Correct, though uninformative: the early endosome (GO:0005769) annotation carried by this gene says where TBCK actually acts. |
| GO:0005769 early endosome | IEA GO_REF:0000044 | ACCEPT | Summary: The compartment at which the RAB5-GAP activity and FERRY complex assembly take place. Reason: Core location, consistent with both the FERRY complex work and the RAB5-GAP work. |
| GO:0005819 spindle | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Derived from the UniProt subcellular-location statement, which rests on a single 2014 immunofluorescence study of a long TBCK isoform. Reason: Retained because the underlying observation is experimental and I have not read its full text, but it is a single unreplicated report, is not connected to the endosomal function established since, and describes a cell-cycle-restricted redistribution rather than the site of TBCK's molecular activity. |
| GO:0030496 midbody | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Derived from the same single 2014 study of a long TBCK isoform, reporting midbody accumulation at the end of mitosis. Reason: Same reasoning as the spindle annotation: experimental, unreplicated, cell-cycle-restricted and disconnected from the endosomal GAP function. |
| GO:0005515 protein binding | IPI PMID:26496610 A human interactome in three quantitative dimensions organiz... | REMOVE | Summary: Bare protein binding from a quantitative interactome survey. Reason: Carries no functional information. The informative content of TBCK's interactions is captured by FERRY complex membership and by the GTPase activator activity term. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:37267906 Structural basis of mRNA binding by the human FERRY Rab5 eff... | REMOVE | Summary: Bare protein binding recorded from the FERRY complex structural study. Reason: Carries no functional information; the underlying finding (TBCK is a FERRY subunit) is better expressed as complex membership than as protein binding. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005769 early endosome | NAS PMID:37267905 The Rab5 effector FERRY links early endosomes with mRNA loca... | ACCEPT | Summary: Author statement from the FERRY complex paper placing the complex, and TBCK with it, on early endosomes. Reason: Core location, independently supported by the RAB5-GAP work. Supporting Evidence: PMID:37267905 we discovered a Rab5 effector, the five-subunit endosomal Rab5 and RNA/ribosome intermediary (FERRY) complex, that recruits mRNAs and ribosomes to early endosomes through direct mRNA-interaction |
| GO:1904580 regulation of intracellular mRNA localization | NAS PMID:37267905 The Rab5 effector FERRY links early endosomes with mRNA loca... | KEEP AS NON CORE | Summary: A genuine process of the FERRY complex, in which TBCK is a subunit, but not an activity of TBCK itself - the mRNA contacts are made by other subunits. Reason: TBCK participates in this process by virtue of complex membership rather than by acting on mRNA. Purified TBCK carries no co-purifying nucleic acid, and the FERRY authors place the mRNA-binding function on other subunits. Retained as a real complex-level involvement, demoted from core because TBCK's own contribution to the complex is now understood to be Rab-directed GAP activity. Supporting Evidence: PMID:42546825 This is consistent with the observation that TBCK does not directly participate in mRNA interactions within the FERRY complex |
| GO:0005737 cytoplasm | EXP PMID:23977024 TBCK influences cell proliferation, cell size and mTOR signa... | ACCEPT | Summary: Experimental cytoplasmic localisation. Reason: Correct, though uninformative: the early endosome (GO:0005769) annotation carried by this gene says where TBCK actually acts. |
| GO:0005737 cytoplasm | IDA PMID:24576458 A long type of TBCK is a novel cytoplasmic and mitotic appar... | ACCEPT | Summary: Experimental cytoplasmic localisation during interphase. Reason: Correct, though uninformative: the early endosome (GO:0005769) annotation carried by this gene says where TBCK actually acts. |
| GO:0008283 cell population proliferation | IMP PMID:23977024 TBCK influences cell proliferation, cell size and mTOR signa... | MARK AS OVER ANNOTATED | Summary: Reduced proliferation after siRNA knockdown in cultured cells, several steps downstream of any molecular activity of TBCK. Reason: Not removed - this is an experimental annotation whose full text I have not read, and the phenotype is presumably real. But proliferation is a distal readout of perturbing endosomal maturation, not a process TBCK carries out, and the same laboratory's mTOR mechanism has not reproduced in other models. Flagged as over-annotation rather than deleted. Supporting Evidence: PMID:23977024 Depletion of TBCK significantly inhibits cell proliferation, reduces cell size, and disrupts the organization of actin, but not microtubule. |
| GO:0030036 actin cytoskeleton organization | IMP PMID:23977024 TBCK influences cell proliferation, cell size and mTOR signa... | MARK AS OVER ANNOTATED | Summary: Disorganised actin after TBCK knockdown in cultured cells. Reason: Same single knockdown experiment as the proliferation and mTOR annotations, with no proposed mechanism linking TBCK to actin and no independent replication. Retained but flagged; an indirect consequence of disturbed endosomal traffic is at least as likely as a direct role. Supporting Evidence: PMID:23977024 Depletion of TBCK significantly inhibits cell proliferation, reduces cell size, and disrupts the organization of actin, but not microtubule. |
| GO:0030496 midbody | IDA PMID:24576458 A long type of TBCK is a novel cytoplasmic and mitotic appar... | KEEP AS NON CORE | Summary: Direct immunofluorescence localisation of a long TBCK isoform to the midbody at the end of mitosis. Reason: Experimental and therefore retained, but unreplicated, cell-cycle-restricted and disconnected from the endosomal GAP function that the rest of the evidence supports. |
| GO:0032006 regulation of TOR signaling | IMP PMID:23977024 TBCK influences cell proliferation, cell size and mTOR signa... | KEEP AS NON CORE | Summary: Contested. TBCK knockdown was reported to lower mTORC component levels and mTOR pathway activity, but this has not reproduced in human neural progenitors or consistently across patient cells. Reason: GOA effectively carries only one side of this question, so the dispute is recorded here. The original 2013 siRNA study reported reduced mTORC subunit levels and reduced mTOR signalling; a human ReNcell neural-progenitor model found no mTORC1 inhibition in progenitors or neurons and found the effect inconsistent across eight patient-derived lines, and the 2026 literature describes TBCK's primary molecular function as having been controversial precisely because of the mTOR claim. Not removed - it is an experimental annotation and the effect may be real in some cell types - but demoted from core, since the current mechanistic account routes the phenotype through RAB5-dependent endolysosomal failure rather than through direct mTOR regulation. Supporting Evidence: PMID:23977024 Knockdown of TBCK induces a significant decrease in the protein levels of components of mTOR complex (mTORC), and suppresses the activity of mTOR signaling PMID:39553985 These data showed that loss of TBCK did not inhibit mTORC1 activity in neither NPC nor neurons. PMID:39553985 mTORC1 inhibition is inconsistent across different patients and cell types PMID:41789809 While the TBCK gene has been implicated in MTOR signaling, its primary molecular function has remained controversial. |
| GO:0072686 mitotic spindle | IDA PMID:24576458 A long type of TBCK is a novel cytoplasmic and mitotic appar... | KEEP AS NON CORE | Summary: Direct immunofluorescence localisation of a long TBCK isoform to the metaphase spindle. Reason: Experimental and therefore retained, but unreplicated and unconnected to the endosomal function; not where TBCK's established molecular activity is exercised. |
| GO:0004672 protein kinase activity | NAS PMID:12471243 The protein kinase complement of the human genome. | REMOVE | Summary: The kinase assignment that the gene symbol asserts. Its evidence is an author statement traced to the human kinome census - an inventory of kinase-fold genes, not an assay - and direct biochemistry on the purified human protein now excludes catalysis. Reason: GOA already carries both sides of this question from the same reference: protein kinase activity (NAS) alongside NOT involved_in protein phosphorylation (NAS). The contradiction resolves against the kinase call. The kinome census enumerated 518 putative kinase genes from sequence, and UniProt annotates the TBCK kinase domain as predicted catalytically inactive. In 2026, purified full-length human TBCK was shown to lack the glycine-rich loop, VAIK, HRD and DFG motifs, to bind no nucleotide, and to hydrolyse neither ATP nor GTP, placing it in class I - the pseudokinase class with no retained ligand binding. No substrate, no phosphotransfer and no kinase assay supporting TBCK has been reported in over twenty years. Caveat, recorded rather than hidden: the biochemistry is from a single laboratory in a methods-oriented journal, and PMID:41958985 is that group's own preprint of the same work, not independent replication. REMOVE is used because this is a non-experimental author statement contradicted by direct measurement, not an experimental annotation being second-guessed. Supporting Evidence: PMID:12471243 We identify 518 putative protein kinase genes, of which 71 have not previously been reported or described as kinases PMID:42546825 Biochemical and biophysical analyses reveal that the catalytically inactive pseudokinase domain of TBCK lacks nucleotide binding, consistent with the absence of the canonical VAIK, HRD, and DFG motifs required for catalysis. PMID:42546825 This indicates that the TBCK pseudokinase domain exhibits no detectable ATPase activity under these conditions |
| GO:0005524 ATP binding | NAS PMID:12471243 The protein kinase complement of the human genome. | REMOVE | Summary: Author statement accompanying the kinome-census kinase call. Purified full-length human TBCK shows no ATP-dependent stabilisation at 1, 5 or 10 mM ATP. Reason: Same basis as the InterPro ATP-binding row and the protein kinase activity row: a sequence-derived assertion of a nucleotide pocket that the protein does not have. Class I pseudokinases are defined by the loss of nucleotide and divalent-cation binding, and TBCK meets that definition experimentally. Supporting Evidence: PMID:42546825 These results, therefore, indicate that none of the tested nucleotides measurably stabilize or destabilize the global thermal stability of TBCK under the conditions examined, consistent with TBCK functioning as a class I pseudokinase. PMID:42546825 Similarly, no GTP hydrolysis was observed in the assay when ATP was replaced with GTP, indicating this is not specific to ATP. |
| GO:0006468 protein phosphorylation | NAS NOT PMID:12471243 The protein kinase complement of the human genome. | ACCEPT | Summary: A NOT annotation: GOA asserts that TBCK is not involved in protein phosphorylation. Read the qualifier - this is the negative of the term, and it is the assertion that the 2026 biochemistry supports. Reason: Correct, and now experimentally grounded rather than merely predicted. Accepting this NOT row while removing the positive protein kinase activity row from the same reference resolves the internal contradiction in GOA in the direction the evidence points. Keeping the explicit negative annotation is valuable: it blocks re-propagation of the kinase call from the gene symbol and from the kinase-fold signature. Supporting Evidence: PMID:42546825 This indicates that the TBCK pseudokinase domain exhibits no detectable ATPase activity under these conditions PMID:42546825 Biochemical and biophysical analyses reveal that the catalytically inactive pseudokinase domain of TBCK lacks nucleotide binding, consistent with the absence of the canonical VAIK, HRD, and DFG motifs required for catalysis. |
| GO:0090630 activation of GTPase activity | IMP PMID:41789809 The TBCK-PPP1R21-FERRY3/C12orf4 complex: a RAB5-GAP brake es... | NEW | Summary: Proposed addition. The direct biological-process consequence of TBCK's GAP activity: stimulation of GTP hydrolysis on RAB5, missing from GOA, which records only the molecular function. Reason: GOA carries GTPase activator activity (IBA) but no corresponding process term. Loss of TBCK, or mutation of the TBC catalytic residues, leaves RAB5 constitutively GTP-bound, which is the phenotype of losing GTPase activation. Supporting Evidence: PMID:41789809 TBCK deficiency or missense mutations of its key residues in the RABGAP-TBC domain lead to constitutive RAB5 hyperactivation, which blocks the transition from early to late endosomes and results in the formation of massively enlarged RAB5-positive endosomes. |
| GO:0045022 early endosome to late endosome transport | IMP PMID:41789809 The TBCK-PPP1R21-FERRY3/C12orf4 complex: a RAB5-GAP brake es... | NEW | Summary: Proposed addition. TBCK loss blocks the conversion of early endosomes to late endosomes, the cellular process its RAB5-GAP activity serves. Reason: This is the process-level correlate of the RAB5-GAP function and the point at which the molecular activity connects to the endolysosomal and autophagic failure seen in TBCK encephaloneuronopathy. Caveat: the underlying primary study (PMID:41207833) was not retrievable, so this rests on the same authors' summary. Supporting Evidence: PMID:41789809 TBCK deficiency or missense mutations of its key residues in the RABGAP-TBC domain lead to constitutive RAB5 hyperactivation, which blocks the transition from early to late endosomes and results in the formation of massively enlarged RAB5-positive endosomes. |
| GO:0160271 FERRY complex | IDA PMID:37267905 The Rab5 effector FERRY links early endosomes with mRNA loca... | NEW | Summary: Proposed addition. TBCK is one of the five subunits of the FERRY complex; GOA records the interaction only as bare protein binding. Reason: Complex membership is the informative form of the two protein binding IPI rows this gene carries, and UniProt already records the five-subunit stoichiometry (TBCK, PPP1R21, FERRY3, CRYZL1, GATD1 at 1:2:1:2:4). Supporting Evidence: PMID:37267905 we discovered a Rab5 effector, the five-subunit endosomal Rab5 and RNA/ribosome intermediary (FERRY) complex, that recruits mRNAs and ribosomes to early endosomes through direct mRNA-interaction |
Loading supporting content…
Download this section (compressed HTML)Q: Should GO:0004672 protein kinase activity and GO:0005524 ATP binding be retired for TBCK on the strength of a single laboratory's in vitro characterisation of the recombinant protein, or should they be held pending independent replication? The result agrees with sequence analysis and with UniProt's own inactivity prediction, but the peer-reviewed paper and the bioRxiv preprint are the same work by the same authors.
Suggested experts: Wenjie Zhu, Shivangi Maurya
Q: Is the mTOR phenotype of TBCK loss a direct regulatory role or a downstream consequence of blocked endosome maturation? The original knockdown study reported reduced mTORC subunit levels, a human neural-progenitor model found no mTORC1 inhibition, and the RAB5-GAP account predicts an indirect, cell-type-dependent effect.
Suggested experts: Elizabeth J K Bhoj, Chenji Wang
Q: Does the pseudokinase domain retain a ligand or a partner, and does it regulate the adjacent TBC domain? Nucleotide binding is excluded, but the AlphaFold model retains a cavity and disease alleles fall outside the TBC domain.
Suggested experts: Wenjie Zhu
Q: Within the FERRY complex, is the RAB5-GAP activity of the TBCK-PPP1R21-FERRY3 heterotrimer compatible with FERRY acting as a RAB5A effector, or do these represent two distinct assemblies with opposite effects on RAB5? A single protein that both binds active RAB5 as part of an effector and stimulates its hydrolysis needs a timing model.
Suggested experts: Marino Zerial, Jan S Schuhmacher
Experiment: Reconstitute RAB5 GAP assays with purified full-length TBCK, the isolated TBC domain, and pseudokinase-domain-deleted protein, each with and without PPP1R21 and FERRY3, measuring single-turnover GTP hydrolysis on RAB5. Compare specific activities and test whether the pseudokinase domain alters the requirement for the partner subunits.
Hypothesis: The pseudokinase domain is an autoinhibitory or targeting module for the TBC domain rather than a vestigial remnant.
Type: In vitro GAP assay with domain-deletion series
Experiment: In isogenic TBCK-null and control lines of several types (HeLa, neural progenitors, patient fibroblasts), measure mTORC1 substrate phosphorylation alongside RAB5-GTP loading and endosome morphology in the same cells, and test whether restoring RAB5 turnover with an orthogonal RAB5 GAP, or with a dominant-negative RAB5, rescues the mTOR readout without restoring TBCK.
Hypothesis: The mTOR signalling changes in TBCK-deficient cells are secondary to blocked early-to-late endosome conversion rather than to a direct action of TBCK on mTORC1.
Type: Isogenic knockout epistasis with pathway-level readouts
Experiment: Perform quantitative phosphoproteomics on isogenic TBCK-null versus control cells, and an analog-sensitive or crosslinking-based substrate-trapping experiment with the kinase-fold domain, to confirm the absence of a phosphotransfer role in a cellular context rather than only in vitro.
Hypothesis: TBCK has no protein kinase substrate in cells.
Type: Quantitative phosphoproteomics plus substrate trapping
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)