coWts (CAOG_00619) is the single Warts/LATS-family (NDR subfamily, AGC group) serine/threonine protein kinase of the filasterean amoeba Capsaspora owczarzaki, a close unicellular relative of animals. The 1205-residue protein has a long, largely disordered N-terminal region, a LATS-type MOB-binding region, a C-terminal kinase domain with the conserved ATP-binding lysine and catalytic aspartate, and an AGC C-terminal extension carrying a predicted hydrophobic-motif phosphorylation site. In Capsaspora it acts in the Hippo pathway upstream of the Yorkie/YAP/TAZ ortholog coYki: cells lacking the coWts kinase domain show increased nuclear accumulation of coYki, more strongly than cells lacking the upstream kinase coHpo, consistent with cytoplasmic retention of coYki by Warts-dependent phosphorylation as in animals, although coWts kinase activity and coYki phosphorylation have not been measured directly. Loss of coWts does not increase proliferation (it slows growth in shaking culture) but produces a subpopulation of elongated cells that undergo a myosin-sensitive contractile behaviour, rescued by a coWts transgene, and within multicellular aggregates causes denser cell packing, reduced aggregate circularity and shorter filopodia. These phenotypes are mimicked by a hyperactive coYki that cannot be phosphorylated at its predicted Warts sites and require coYki's TEAD-binding residue, indicating that coWts shapes cytoskeletal behaviour and aggregate architecture by restraining coYki-dependent transcription.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000166 nucleotide binding | IEA GO_REF:0000104 | KEEP AS NON CORE | Summary: UniRule-based assignment from the protein kinase ATP-binding signature. coWts has a canonical protein kinase domain (residues 711-1045) with the PROSITE ATP-binding motif and the conserved lysine (K740). Reason: Correct but uninformative: nucleotide binding is the parent of the ATP binding annotation already present, and ATP binding is itself only a component of the kinase activity. Kept as a true, non-core statement; no experimental nucleotide binding data exist. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PROSITE; PS00107; PROTEIN_KINASE_ATP; 1. |
| GO:0004672 protein kinase activity | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: InterPro2GO from the Ser/Thr kinase active-site signature (IPR008271). The catalytic loop aspartate (D834, proton acceptor) is annotated as present. Reason: Correct, but the more specific protein serine/threonine kinase activity term is also present and is the better description of a Warts/LATS AGC kinase. Kept as a true, less specific parent. Activity has not been assayed for coWts. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PROSITE; PS00108; PROTEIN_KINASE_ST; 1. |
| GO:0004674 protein serine/threonine kinase activity | IEA GO_REF:0000120 | ACCEPT | Summary: Combined IEA (ARBA, InterPro AGC C-terminal domain, EC 2.7.11.1, PANTHER, UniRule). coWts is the Capsaspora Warts/LATS ortholog: it has a LATS-type MOB-binding region (CDD cd21774), a kinase domain with intact ATP-binding lysine and catalytic aspartate, and an AGC C-terminal extension. In animals Warts/LATS phosphorylates Yorkie/YAP/TAZ on HXRXXS motifs; coYki carries four such motifs, and the non-phosphorylatable coYki 4SA mutant is nuclear and phenocopies the coWts knockout [PMID:35659869, PMID:38517944]. Reason: This is the core molecular activity of a Warts/LATS kinase and nothing in the sequence argues against it. The PANTHER component of this row comes from a node in PTHR22988 (Rho-effector kinase family) rather than the LATS clade, but serine/threonine kinase activity is shared by both families, so the misplacement does not affect this term. Note that no in vitro kinase assay or phosphorylation readout (phospho-specific antibody, Phos-tag) has been reported for coWts; the activity is inferred from domain conservation and the 4SA genetic mimicry, not measured. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt CDD; cd21774; MobB_LATS; 1. PMID:35659869 In mammals and in Drosophila, YAP/TAZ/Yorkie is regulated by cytoplasmic sequestration in response to phosphorylation of HXRXXS motifs by the LATS/Warts kinases (Zheng and Pan, 2019). PMID:38517944 These results suggest that the phenotypes observed in coHpo and coWts mutants are due to coYki activation resulting from a loss of upstream kinase signaling. |
| GO:0005524 ATP binding | IEA GO_REF:0000120 | ACCEPT | Summary: Domain-based assignment from the protein kinase domain and its ATP-binding signature (K740). Reason: Kinase domain with conserved glycine-rich loop/ATP-binding lysine; ATP binding is required for the predicted catalytic activity. Accepted as a sound domain-level inference, although ATP binding has not been measured. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt InterPro; IPR017441; Protein_kinase_ATP_BS. |
| GO:0005737 cytoplasm | IEA GO_REF:0000120 | ACCEPT | Summary: Combined ARBA/SubCell/PANTHER inference. The localization of coWts protein has not been examined. Its inferred substrate, coYki, is retained in the cytoplasm in wild-type cells and becomes nuclear when coWts is lost [PMID:38517944], consistent with coWts acting in the cytoplasm as Warts/LATS does in animals. Reason: Cytoplasm is the expected compartment for a Warts/LATS kinase, which retains its substrate in the cytoplasm. The term is broad enough to be safe without direct localization data. Accepted as a homology-based location; it should be confirmed by tagging coWts. Supporting Evidence: PMID:38517944 Together, these results show that the Capsaspora Hippo kinase cascade regulates coYki by cytoplasmic sequestration, indicating a premetazoan origin of this regulatory activity of the Hippo pathway. |
| GO:0005856 cytoskeleton | IEA GO_REF:0000118 | MARK AS OVER ANNOTATED | Summary: TreeGrafter transfer from PANTHER node PTN001122925. TreeGrafter places coWts in PTHR22988:SF71 "CITRON RHO-INTERACTING KINASE" (family PTHR22988, whose human members are ROCK1/2, MRCKs and CIT), not in the LATS clade where human LATS1/2 sit (PTHR24356). No localization of coWts has been reported. Reason: The cytoskeletal location is characteristic of the Rho-effector kinases of PTHR22988, and reaches coWts because of the family placement rather than anything known about Warts/LATS. coWts's MOB-binding region and C-terminal kinase domain identify it as a LATS kinase. Animal LATS kinases are cytoplasmic with some cortical/centrosomal pools, so an association with cytoskeletal structures cannot be ruled out, but there is no evidence for it here. Not removed outright because the term is not demonstrably false. Propagation Review Root cause: PROPAGATION BAD Failure modes: WRONG ORTHOLOG OR PARALOG Sources checked: PANTHER:PTN001122925 SUPPORTS SOURCE BUT NOT TARGET TreeGrafter node in PTHR22988 (Rho-effector kinases); its cytoskeleton and actomyosin terms fit ROCK/MRCK/citron, while coWts is a Warts/LATS kinase whose human orthologs LATS1/2 are in PTHR24356 and receive IBAs from LATS node PTN002390470. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PANTHER; PTHR22988:SF71; CITRON RHO-INTERACTING KINASE; 1. file:CAPO3/coWts/coWts-uniprot.txt CDD; cd21774; MobB_LATS; 1. |
| GO:0007010 cytoskeleton organization | IEA GO_REF:0000117 | MARK AS OVER ANNOTATED | Summary: ARBA machine-learning rule. In Capsaspora, coWts knockout cells show a subpopulation of elongated, contractile cells (rescued by a coWts transgene) and shorter filopodia within aggregates [PMID:38517944]. These effects are phenocopied by hyperactive coYki and require coYki transcriptional activity. Reason: coWts influences cytoskeletal behaviour, but indirectly: the phenotypes are attributed to derepressed coYki-dependent transcription, not to coWts acting on cytoskeletal structures. By the participation test, coWts does not do the work of assembling or arranging the cytoskeleton. Process terms describing the signalling role (hippo signaling) capture its function better. The F-actin, blebbistatin and latrunculin analyses were also performed in the coHpo mutant rather than coWts. Supporting Evidence: PMID:38517944 This result indicates that these phenotypes are mediated by the transcriptional activity of coYki. PMID:38517944 As contractile cells in the coHpo mutant background tended to show a more elongated morphology than the coWts mutant, we focused on the coHpo mutant for further analysis. |
| GO:0016301 kinase activity | IEA GO_REF:0000104 | KEEP AS NON CORE | Summary: UniRule assignment of the general kinase activity parent term. Reason: True but redundant with the more specific protein serine/threonine kinase activity annotation. Kept as a non-core parent. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PROSITE; PS50011; PROTEIN_KINASE_DOM; 1. |
| GO:0016740 transferase activity | IEA GO_REF:0000104 | KEEP AS NON CORE | Summary: UniRule assignment of the very general transferase activity term. Reason: True (a kinase transfers phosphate) but far too general to be informative. Kept as a non-core parent of protein serine/threonine kinase activity. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PROSITE; PS50011; PROTEIN_KINASE_DOM; 1. |
| GO:0031032 actomyosin structure organization | IEA GO_REF:0000118 | REMOVE | Summary: TreeGrafter transfer from PANTHER node PTN001122925 in PTHR22988, the ROCK/MRCK/citron Rho-effector kinase family, whose members directly phosphorylate myosin light chain and regulators of actomyosin. coWts is a Warts/LATS kinase (MOB-binding region, C-terminal kinase domain) that falls into this family's HMM while its human orthologs LATS1/2 are classified in PTHR24356. The coWts knockout does show a myosin-sensitive contractile cell behaviour and denser aggregates [PMID:38517944], but actomyosin structures were imaged and perturbed only in coHpo mutants and wild type, and the phenotypes are attributed to coYki transcription. Reason: The term comes from a family-level placement with the Rho-effector kinases, which do organize actomyosin directly; it is not inherited from the Warts/LATS lineage. The Capsaspora data do not rescue it: coWts loss increases a contractile behaviour through coYki-dependent transcription (phenocopied by coYki 4SA, abolished by the TEAD-binding F123A mutation), there was no epistasis test, and the authors state that the role of myosin in these phenotypes remains unclear. coWts therefore acts upstream of, not in, actomyosin structure organization. This is the Track C pattern of a family-specific cytoskeletal process term transferred to a unicellular protein through a misassigned family. Propagation Review Root cause: PROPAGATION BAD Failure modes: WRONG ORTHOLOG OR PARALOG ROLE CONFLATION Sources checked: PANTHER:PTN001122925 SUPPORTS SOURCE BUT NOT TARGET PTHR22988 node (ROCK/MRCK/citron family); actomyosin organization is a direct activity of those Rho-effector kinases, but coWts belongs to the Warts/LATS clade, which acts on actomyosin only indirectly via Yorkie. Supporting Evidence: file:CAPO3/coWts/coWts-uniprot.txt PANTHER; PTHR22988; MYOTONIC DYSTROPHY S/T KINASE-RELATED; 1. PMID:38517944 This result indicates that these phenotypes are mediated by the transcriptional activity of coYki. PMID:38517944 Thus the specific role of myosin activity in the phenotypes observed in Hippo pathway mutants in Capsaspora remains unclear. |
| GO:0071944 cell periphery | IEA GO_REF:0000117 | UNDECIDED | Summary: ARBA machine-learning rule. In animals Warts/LATS is recruited to the plasma membrane by Mats/MOB1 and Merlin/Kibra during activation; Capsaspora encodes Kibra and Merlin orthologs. The localization of coWts itself has not been examined. Reason: A membrane-proximal pool is plausible for a Warts/LATS kinase, but there are no Capsaspora data for or against it, and the rule-based assignment is not specific to the LATS clade. Neither acceptance nor removal is justified without localization of tagged coWts. Supporting Evidence: PMID:38517944 The Capsaspora genome encodes all components of the core Hippo pathway as well as some conserved upstream Hippo pathway regulators such as Kibra and NF2/Merlin |
| GO:1900181 negative regulation of protein localization to nucleus | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA rule. Directly supported in Capsaspora: in coWts knockout cells mScarlet-coYki is nuclear in most cells, whereas it is predominantly cytoplasmic in wild type; the effect is stronger than in coHpo knockouts [PMID:38517944]. Reason: Restraining nuclear entry of Yorkie/YAP/TAZ is the canonical output of Warts/LATS and is shown genetically for coWts by the knockout (IMP-grade evidence from PMID:38517944, recorded here rather than as a duplicate row). Caveat: the mechanism (phosphorylation of coYki) is inferred from the 4SA mutant, not measured. Supporting Evidence: PMID:38517944 In contrast, the majority of coHpo-/- and coWts-/- cells showed nuclear levels of mScarlet-coYki greater than or equal to that seen in the cytoplasm (Figure 1C and D). |
| GO:0035329 hippo signaling | IMP PMID:38517944 The Hippo kinase cascade regulates a contractile cell behavi... | NEW | Summary: coWts is the Capsaspora Warts/LATS kinase of the Hippo cascade. Its knockout releases coYki to the nucleus, and its cytoskeletal and aggregate phenotypes are phenocopied by coYki 4SA (predicted Warts sites mutated) in a TEAD-binding-dependent manner, matching the canonical Hpo-Wts-Yki-Sd output. Reason: Participation test: Warts/LATS performs a step of the cascade (phosphorylation of Yorkie), so it is a participant rather than a regulated target. Comparator check: Drosophila wts and human LATS1/LATS2 carry GO:0035329 (experimental and IBA from LATS node PTN002390470). coWts received no hippo signaling IEA/IBA only because TreeGrafter placed it in PTHR22988 rather than the LATS clade. The IMP rests on the pathway readout (coYki nuclear localization) and the 4SA phenocopy. Limits: coHpo-to-coWts and coWts-to-coYki phosphorylation were not shown, and coWts may also have coHpo-independent activity. Supporting Evidence: PMID:38517944 Together, these results show that the Capsaspora Hippo kinase cascade regulates coYki by cytoplasmic sequestration, indicating a premetazoan origin of this regulatory activity of the Hippo pathway. PMID:38517944 These results suggest that the phenotypes observed in coHpo and coWts mutants are due to coYki activation resulting from a loss of upstream kinase signaling. PMID:35659869 In contrast, the majority of cells transfected with mScarlet-coYki 4SA showed uniform mScarlet signal throughout the cell or enriched mScarlet signal in the nucleus (Figure 5C and F). |
Loading supporting content…
Download this section (compressed HTML)Q: Does coWts phosphorylate coYki on its HXRXXS motifs in vitro and in cells, and does coHpo phosphorylate coWts on its activation-loop and hydrophobic-motif sites?
Q: Why is the coYki nuclear-localization phenotype stronger in coWts than in coHpo mutants - which kinases (e.g. other STE20/MAP4K-like kinases) activate coWts independently of coHpo?
Q: Where is coWts located in Capsaspora cells (cytoplasm, cortex, filopodia, adhesion foci), and is it recruited to membranes by Mats/MOB, Kibra or Merlin orthologs?
Q: Should PANTHER place Warts/LATS orthologs such as coWts and Drosophila wts in PTHR24356 with human LATS1/2 rather than in PTHR22988 (Rho-effector kinases), so that they inherit the LATS-node IBAs (hippo signaling) instead of the ROCK/citron actomyosin terms?
Q: Which coYki target genes (e.g. the actin-binding and laminin G-domain genes shared among the coHpo, coWts and coYki RNA-seq sets) mediate the contractile behaviour and cell-packing phenotype?
Q: What explains the slower growth of coWts mutants in shaking culture - a coYki-independent function of coWts, or a viability effect of coYki hyperactivity?
Experiment: In vitro kinase assay with immunopurified wild-type versus kinase-dead (K740A or D834A) coWts on recombinant coYki and coYki 4SA, plus Phos-tag or phospho-specific immunoblots of coYki in wild-type and coWts-/- Capsaspora cells.
Hypothesis: coWts is an active kinase that phosphorylates coYki at its predicted Warts sites.
Type: biochemical kinase assay
Experiment: Complement coWts-/- cells with wild-type versus kinase-dead coWts and score coYki nuclear localization, elongated-cell frequency and aggregate cell density.
Hypothesis: Kinase activity, not only the protein scaffold, is required for coWts function.
Type: genetic complementation
Experiment: Construct coWts coYki double mutants (once sequential marker recycling is available) or deplete coYki in coWts-/- cells and test whether contractile behaviour, cell packing and the filopodial defect are suppressed.
Hypothesis: coWts acts upstream of coYki for all scored phenotypes.
Type: genetic epistasis
Experiment: Express fluorescently tagged coWts from its endogenous locus or at low levels and image together with Lifeact and membrane (NMM) markers in adherent cells and aggregates.
Hypothesis: coWts localizes to the cell cortex or adhesion foci.
Type: localization / imaging
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)