Capsl encodes calcyphosin-like protein, a small (208 aa) cytoplasmic protein made of four EF-hand domains arranged as two EF-hand pairs, in the calcyphosin family (PANTHER PTHR34524, subfamily SF5, calcyphosin-like protein). EF-hands 1 and 2 have calcium-coordinating loops, while EF-hands 3 and 4 lack the glutamate at loop position 12 and are probably not calcium-binding; calcium binding has not been measured for the mouse protein. Its molecular activity is unknown. The only in vivo experimental data in mouse come from an inducible endothelial-specific knockout (Pdgfb-iCreER). In this model, postnatal retinal vascularization is delayed, with slower radial and deep-plexus growth, fewer tip cells and filopodia, less endothelial proliferation, more vessel regression, loss of tip-cell polarity and delayed hyaloid regression. These defects are endothelial-autonomous and transient, and resemble familial exudative vitreoretinopathy, for which human CAPSL is a candidate gene. In knockdown experiments in human retinal endothelial cells, CAPSL loss lowered MYC protein, but not MYC mRNA, and reduced MYC target-gene programmes. The molecular link between CAPSL and MYC is unknown. In mouse 3T3-L1 preadipocytes, Capsl knockout enhanced adipogenic differentiation and altered autophagy markers. In bovine tracheal cilia, cryo-electron tomography (a preprint) identified the ortholog as a lumenal microtubule inner protein of the transition-zone doublet. This ciliary role is inferred for mouse Capsl by orthology within the same PANTHER subfamily and has not been tested in mouse.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0001525 angiogenesis | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | MODIFY | Summary: MGI IMP based on the inducible endothelial-specific Capsl knockout (iECKO). At P5, iECKO retinas show reduced radial vascular outgrowth, vessel density and branching, and fewer tip cells and filopodia at the angiogenic front. Reason: Endothelial Capsl is required for retinal angiogenesis, and the acts_upstream_of_or_within qualifier fits a requirement whose mechanism is unknown. However, the data are more specific than "angiogenesis". The phenotypes are features of sprouting angiogenesis: tip-cell number, filopodia extension, tip-cell polarity, and stalk-cell proliferation at the growing front. The replacement is GO:0002040 sprouting angiogenesis, an is_a descendant of angiogenesis, with the same acts_upstream_of_or_within qualifier. No molecular role of Capsl within the sprouting process has been shown, so this is a requirement claim, not a claim that Capsl performs a step. Proposed replacements: sprouting angiogenesis Supporting Evidence: PMID:39264149 the radial vascular growth, vessel density, and vascular branching were dramatically reduced in P5 iECKO retina relative to Ctrl mice PMID:39264149 The number of retinal vascular tip cells, which leads EC migration toward high vascular endothelial growth factor (VEGF) by extending filopodia at the angiogenic front (Rattner et al., 2019), was considerably decreased and sparsely distributed in iECKO retinae |
| GO:0001570 vasculogenesis | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | MARK AS OVER ANNOTATED | Summary: MGI IMP from the same endothelial-specific knockout study. The knockout was induced at P1-P3, after the embryonic vasculature had formed, and the phenotypes were scored in the postnatal retina. Reason: GO defines vasculogenesis as de novo vessel formation from angioblasts. The paper does not assay de novo vessel formation, and it describes the retinal phenotype as impaired angiogenesis (expansion of an existing plexus by sprouting). Tamoxifen was given at P1-P3, after embryonic vasculogenesis is complete. The paper reports no embryonic vascular phenotype and says iECKO mice look normal overall. The term therefore goes beyond the evidence. The supported claim is captured by sprouting angiogenesis and retina vasculature development. The row is not REMOVE, because the experimental data are sound, and the curator may have counted retinal vascular plexus formation as vasculogenesis. Supporting Evidence: PMID:39264149 were induced by consecutive intraperitoneal injection of 50 ΞΌg tamoxifen per pup at postnatal day 1 (P1) to P3 before sacrifice PMID:39264149 iECKO mice showed indistinguishable overall appearance compared to their Ctrl littermates. |
| GO:0001935 endothelial cell proliferation | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | ACCEPT | Summary: In vivo EdU/ERG labelling of P5 iECKO retinas showed strongly reduced endothelial proliferation. CAPSL knockdown in human retinal endothelial cells likewise reduced EdU incorporation and caused G0/G1 arrest. Reason: This is a direct in vivo readout in mouse, with the gene deleted only in endothelial cells (cell-autonomous context). The acts_upstream_of_or_within qualifier is appropriate, because the mechanism is probably indirect: CAPSL loss lowers MYC protein and MYC/E2F target programmes in human endothelial cells. The term is as specific as the evidence allows. "Blood vessel endothelial cell proliferation involved in sprouting angiogenesis" would also fit, but adds little. This is the best-supported in-species process for Capsl. Supporting Evidence: PMID:39264149 The results showed a strongly reduced proliferation of Capsl-defective ECs compared to that of Ctrl (Figure 4A). |
| GO:0003674 molecular_function | ND GO_REF:0000015 | ACCEPT | Summary: Root-term ND annotation: the curator found no experimental data on the molecular function of mouse Capsl. Reason: This is still correct for experimental evidence in mouse. No biochemical activity of mouse Capsl has been measured. Calcium ion binding is predicted from EF-hand motifs only (the electronic row below). Microtubule binding and stabilization were shown only for the bovine ortholog (native structure) and in vitro, in a preprint. No GO_REF:0000015 literature-search ND passage is quoted here, because the UniProt record has no function comment for this protein. The ND stands until a mouse activity is measured. The gene's notes survey of what is not experimental in mouse finds the mechanistic data only in human cells and calcium binding only predicted. Supporting Evidence: file:mouse/Capsl/Capsl-notes.md All mechanistic data come from human HRECs, not mouse ... Calcium binding is predicted only. |
| GO:0005509 calcium ion binding | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro2GO mapping from the EF-hand domain (IPR002048). Mouse Capsl has four PROSITE EF-hand domains. Calcium-binding residues are annotated in loops 1 (residues 52-63) and 2 (residues 88-99). Reason: Two EF-hand loops are predicted to coordinate Ca2+. Loop 2 (DRDGSGTIDFNE) is fully canonical. Loop 1 (DDNNNRTLDFKE) keeps the coordinating Asp/Asn residues and the position-12 Glu. Loops 3 and 4 lack the position-12 Glu. The calcyphosin family fold, two EF-hand pairs, is calcium-loaded in the crystal structure of the paralog calcyphosin (CAPS, a different subfamily of the same PANTHER family). Calcium binding is therefore well supported as a family-level property, but it has not been measured for mouse or human CAPSL. It is accepted as the only molecular function with domain-level support, and it remains inferred. Supporting Evidence: file:mouse/Capsl/Capsl-uniprot.txt /ligand="Ca(2+)" PMID:18775726 The crystal structure of Ca(2+)-loaded calcyphosine was determined up to 2.65 A resolution and reveals a protein containing two pairs of Ca(2+)-binding EF-hand motifs. |
| GO:0005575 cellular_component | ND GO_REF:0000015 | ACCEPT | Summary: Root-term ND annotation: no experimental data on the location of mouse Capsl. Reason: This is still correct for experimental evidence in mouse. The cytoplasmic location is a UniProt curator inference (ECO:0000305). The cytoplasm/nucleus statement in the FEVR study cites the Human Protein Atlas, and its immunostaining used overexpressed human CAPSL in 293T cells. The transition-zone localization comes from bovine cilia. None of these is a mouse localization experiment. The gene's own literature record (notes) finds no mouse localization experiment. Supporting Evidence: file:mouse/Capsl/Capsl-notes.md There is no mouse localization experiment. |
| GO:0005737 cytoplasm | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: UniProt subcellular-location mapping. The UniProt entry gives Cytoplasm with ECO:0000305 (curator inference). Reason: Mouse Capsl is a soluble EF-hand protein with no signal peptide or transmembrane segment, so a cytoplasmic location is plausible. Human CAPSL is reported in cytoplasm and nucleus (Human Protein Atlas, cited in PMID:39264149). The most specific localization known for the ortholog is the lumen of the ciliary transition-zone doublet microtubules (bovine cryo-ET preprint). Here it is not proposed as a replacement for mouse, because it has not been tested in mouse. The broad term is kept as non-core. Supporting Evidence: file:mouse/Capsl/Capsl-uniprot.txt SUBCELLULAR LOCATION: Cytoplasm {ECO:0000305}. PMID:39264149 CAPSL is localized in both the cytoplasm and the nucleus of cells (https://www.proteinatlas.org/) |
| GO:0008150 biological_process | ND GO_REF:0000015 | REMOVE | Summary: Root-term ND annotation for biological process. Experimental IMP process annotations from PMID:39264149 now exist for the same gene. Reason: ND means no biological-process data are available. That is no longer true: the 2024 endothelial knockout study provides IMP evidence for endothelial proliferation, endothelial migration and retinal sprouting angiogenesis, which MGI has annotated. The ND row is obsolete and should be retired. Supporting Evidence: PMID:39264149 The results showed a strongly reduced proliferation of Capsl-defective ECs compared to that of Ctrl (Figure 4A). |
| GO:0010467 gene expression | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | MARK AS OVER ANNOTATED | Summary: Based on RT-qPCR showing lower expression of MYC target genes in lung tissue of iECKO mice. The underlying mechanism comes from human endothelial cells, where CAPSL knockdown lowered MYC protein (not mRNA) and MYC/E2F target transcripts. Reason: Lower expression of a set of downstream target genes after knockout is a readout, not evidence that Capsl takes part in gene expression. Under any qualifier, "gene expression" is uninformative. The specific finding, which is human-cell and partly mechanistic, is that CAPSL maintains MYC protein levels post-transcriptionally. This might eventually support a term such as regulation of protein stability or MYC signalling. However, the mechanism (stability, translation, or degradation) has not been determined, and the key experiment was done in human cells. No replacement is proposed. The mouse result is noted under suggested_experiments. Supporting Evidence: PMID:39264149 RT-qPCR analysis was further performed to validate the downregulation fo MYC target genes in the lung tissue of Ctrl and iECKO mice PMID:39264149 Notably, the depletion of CAPSL significantly decreased the protein level of MYC, rather than the mRNA level in HRECs (Figure 7G). |
| GO:0043542 endothelial cell migration | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | ACCEPT | Summary: In iECKO retinas, tip-cell nuclei were more spherical and failed to orient toward the avascular area, an in vivo proxy for migratory polarity. In human retinal endothelial cells, CAPSL knockdown impaired wound closure, Golgi polarization, filopodia and lamellipodia, and lowered CDC42, RAC1, RHOA and MYL9. Reason: The mouse evidence for migration is indirect: tip-cell nuclear shape and orientation, plus reduced vascular progression. The direct migration assays are in human cells, from an ortholog in the same PANTHER subfamily with about 95% identity. Taken together, the data support an acts_upstream_of_or_within role in endothelial migration in mouse. The mechanism is unknown, so the qualifier is appropriate. The term is as specific as warranted. Supporting Evidence: PMID:39264149 Conversely, in the retinae of iECKO mice, the nuclei of tip ECs appeared more spherical and did not orient toward the avascular area (Figure 4C). PMID:39264149 shCAPSL-ECs exhibited fewer filopodia, smaller lamellipodia, and significantly impaired wound closure compared to shCtrl-ECs at the wound scratch edge |
| GO:0060041 retina development in camera-type eye | IMP PMID:39264149 Dysfunction of Calcyphosine-Like gene impairs retinal angiog... | MODIFY | Summary: Based on delayed superficial and deep retinal vascular development and delayed hyaloid regression in iECKO mice. The neural retina, astrocyte template, pericytes, VEGFA gradient and ECM were all normal. Reason: The phenotype is limited to the retinal vasculature. The authors found no defect in astrocytes, pericytes, VEGFA, ECM or the blood-retina barrier, and deleted the gene only in endothelial cells. "Retina development" is too broad and suggests a neural-retina role. The specific term is GO:0061298 retina vasculature development in camera-type eye, which is part of retina development. It keeps the same acts_upstream_of_or_within qualifier. The defect is transient: by P21 the vasculature resembles controls. Proposed replacements: retina vasculature development in camera-type eye Supporting Evidence: PMID:39264149 In conclusion, EC-specific knockout of Capsl may mainly lead to defects in retinal vascular ECs rather than other vascular-associated cells. PMID:39264149 The results revealed a delayed development of deep retinal vessels in iECKO mouse retinae compared to that of Ctrl (Figure 3B). |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Is mouse Capsl a lumenal microtubule inner protein of the ciliary transition zone, as reported for the bovine ortholog? If so, does its endothelial requirement act through the endothelial primary cilium, or through an unrelated cytoplasmic role?
Q: Does Capsl bind calcium with physiologically relevant affinity, and does calcium change its microtubule association?
Q: Should the MGI vasculogenesis IMP (PMID:39264149) be retired, given that deletion was induced at P1-P3 and only postnatal retinal angiogenesis was assessed?
Experiment: Immunolocalize endogenous Capsl, or a knock-in tag, in mouse tracheal multiciliated cells and in retinal endothelial primary cilia, with transition-zone markers (e.g. CEP290, RPGRIP1L), to test the transition-zone localization inferred from the bovine ortholog.
Type: immunofluorescence localization
Experiment: Measure MYC protein and mRNA in sorted retinal endothelial cells from iECKO and control mice, and test whether endothelial Myc overexpression rescues the retinal sprouting defect.
Type: genetic rescue
Experiment: Measure calcium binding of recombinant mouse Capsl (ITC or fluorescence titration), using EF-hand 1 and 2 loop mutants as controls.
Type: biophysical binding assay
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)