TMEM63B

UniProt ID: Q5T3F8
Organism: Homo sapiens
Review Status: COMPLETE
📝 Provide Detailed Feedback

Gene Description

Mechanically activated, Ca2+-permeable cation channel of the TMEM63/OSCA (CSC1) family that also acts as a mechanically activated phospholipid scramblase. TMEM63B is an 11-transmembrane protein that, unusually for this structural superfamily, functions as a monomer with a single highly restricted pore, and it gates according to the force-from-lipids principle: membrane stretch or hypo-osmotic swelling opens a small-conductance, high-threshold cation conductance permeable to Ca2+, Na+, K+ and Mg2+. The same protein translocates phospholipids bidirectionally between bilayer leaflets when membrane curvature and thickness change, dissipating plasma membrane lipid asymmetry and altering the distribution of phosphatidylcholine and sphingomyelin; cryo-EM structures resolve a membrane-spanning lipid pathway that opens in the same groove as the ion pore, and distinct residues form the bottlenecks for ions and for lipids. This dual channel/scramblase behaviour parallels the TMEM16 family, to which TMEM63 proteins are structurally related. TMEM63B is found in the plasma membrane, lysosomal and early endosomal membranes, and the limiting membrane of alveolar lamellar bodies. Physiologically it is required for stretch-evoked pulmonary surfactant and ATP secretion from alveolar type 2 cells, and in rodents it contributes to osmosensing for thirst, to outer hair cell survival and hearing, and to intestinal stem cell homeostasis. Heterozygous de novo missense and in-frame variants clustered around the hydrophobic gate cause developmental and epileptic encephalopathy with progressive neurodegeneration, often with macrocytic anaemia and haemolysis; several of these variants act by converting the protein into a constitutive scramblase rather than by increasing ion leak. Whether the wild-type scramblase activity is deployed physiologically in vivo, as opposed to being measurable in reconstituted and over-expressing systems, is not yet settled.

Proposed New Ontology Terms

phosphatidylcholine scramblase activity

Definition: Catalysis of the movement of phosphatidylcholine from one leaflet of a membrane bilayer to the other, by an ATP-independent mechanism.

Justification: GO has no lipid-species-specific child of GO:0017128 phospholipid scramblase activity. Because of this gap, curators annotating the demonstration that TMEM63B redistributes phosphatidylcholine and sphingomyelin across the plasma membrane reached for GO:0120019 phosphatidylcholine transfer activity and GO:0140338 sphingomyelin transfer activity, which describe soluble intermembrane lipid-transfer proteins that shuttle a lipid through the aqueous phase in a hydrophobic pocket - a mechanism TMEM63B does not use. Species-specific scramblase children would let this evidence be captured without asserting the wrong mechanism.

Parent term: phospholipid scramblase activity

Supporting Evidence:

Existing Annotations Review

GO Term Evidence Action Reason
GO:0005227 calcium-activated cation channel activity
IBA
GO_REF:0000033
MODIFY
Summary: Ca2+-permeable, stretch/osmolarity-gated cation channel annotated with a term that asserts gating by Ca2+ binding.
Reason: The essence is right - TMEM63B is a Ca2+-permeable cation channel - but the term names the wrong gating stimulus. GO:0005227 is defined as a channel that opens 'when a calcium cation has been bound by the channel complex or one of its constituent parts', and every experimental user of the term in GOA is a genuinely Ca2+-gated channel (KCNN4, SLO1/SLO2, TRPM4, TMEM16A). TMEM63B is gated by membrane stretch and hypo-osmolarity, with Ca2+ as a permeant ion. The provenance of the slip is visible: the mouse IDA that seeds the human ISS/IEA rows comes from a paper titled 'Overexpression of Osmosensitive Ca(2+)-Permeable Channel TMEM63B' (PMID:31243992), and the InterPro family behind the IEA is named 'Calcium permeable stress-gated cation channel 1-like' (IPR045122) yet is mapped to the calcium-activated term. Replace with the mechanosensitive cation channel term plus a plain calcium channel term, which together capture what was actually measured without asserting Ca2+ gating.
Propagation Review
Root cause: TERM SCOPING PROBLEM
Failure modes: GRANULARITY MISMATCH
Sources checked:
PANTHER:PTN001404773 · OSCA/TMEM63 ancestral node SUPPORTS SOURCE BUT NOT TARGET
The node correctly captures a conserved stress-gated, Ca2+-permeable cation channel; the GO term chosen for it names Ca2+ as the gating ligand instead.
MGI:MGI:2387609 · Tmem63b (mouse) SUPPORTS SOURCE BUT NOT TARGET
Mouse IDA rests on PMID:31243992, titled 'Overexpression of Osmosensitive Ca(2+)-Permeable Channel TMEM63B', which reports Ca2+ permeability and hyperosmolarity gating, not Ca2+-dependent gating.
UniProtKB:Q5T3F8 · TMEM63B (human, the target itself) SUPPORTS SOURCE BUT NOT TARGET
Expected and legitimate self-inclusion: the human IDA on this term is one of the descendant evidences behind the IBD. It carries the same term-choice problem.
UniProtKB:Q9P1W3 · TMEM63C (human) SUPPORTS SOURCE BUT NOT TARGET
AGI_LocusCode:AT4G04340 · Arabidopsis OSCA SUPPORTS SOURCE BUT NOT TARGET
Plant OSCA proteins are hyperosmolality-gated calcium-permeable channels; the same permeable/activated conflation applies at the plant end of the tree.
AGI_LocusCode:AT4G22120 · Arabidopsis OSCA SUPPORTS SOURCE BUT NOT TARGET
SGD:S000004231 · yeast CSC1-like UNRESOLVED
Not individually inspected.
GO:0005886 plasma membrane
IBA
GO_REF:0000033
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0005227 calcium-activated cation channel activity
IEA
GO_REF:0000120
MODIFY
Summary: Ca2+-permeable, stretch/osmolarity-gated cation channel annotated with a term that asserts gating by Ca2+ binding.
Reason: The essence is right - TMEM63B is a Ca2+-permeable cation channel - but the term names the wrong gating stimulus. GO:0005227 is defined as a channel that opens 'when a calcium cation has been bound by the channel complex or one of its constituent parts', and every experimental user of the term in GOA is a genuinely Ca2+-gated channel (KCNN4, SLO1/SLO2, TRPM4, TMEM16A). TMEM63B is gated by membrane stretch and hypo-osmolarity, with Ca2+ as a permeant ion. The provenance of the slip is visible: the mouse IDA that seeds the human ISS/IEA rows comes from a paper titled 'Overexpression of Osmosensitive Ca(2+)-Permeable Channel TMEM63B' (PMID:31243992), and the InterPro family behind the IEA is named 'Calcium permeable stress-gated cation channel 1-like' (IPR045122) yet is mapped to the calcium-activated term. Replace with the mechanosensitive cation channel term plus a plain calcium channel term, which together capture what was actually measured without asserting Ca2+ gating.
GO:0005765 lysosomal membrane
IEA
GO_REF:0000120
ACCEPT
Summary: TMEM63B also resides in the lysosomal membrane.
Reason: Experimentally supported secondary location; the scramblase work localises TMEM63B to both the plasma membrane and lysosomes.
GO:0005789 endoplasmic reticulum membrane
IEA
GO_REF:0000044
KEEP AS NON CORE
Summary: Endoplasmic reticulum membrane pool, supported by orthology to mouse Tmem63b.
Reason: Plausible and supported experimentally in mouse, but for a plasma-membrane/lysosomal channel-scramblase the ER pool most likely reflects the biosynthetic route rather than a site of action. Retained as a secondary location.
GO:0005886 plasma membrane
IEA
GO_REF:0000120
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0015914 phospholipid transport
IEA
GO_REF:0000108
MODIFY
Summary: Generic phospholipid transport, derived electronically from the phosphatidylcholine transfer activity annotation.
Reason: True but uninformative, and its electronic basis (GO:0120019) is itself the wrong mechanism class. The specific process TMEM63B carries out is plasma membrane phospholipid scrambling, a descendant of this term which the gene already carries.
GO:0016020 membrane
IEA
GO_REF:0000002
MODIFY
Summary: Root-level membrane term from an InterPro domain match.
Reason: Correct but carries no information beyond 'is a membrane protein'. The specific membranes are known and already annotated.
Proposed replacements: plasma membrane
GO:0017121 plasma membrane phospholipid scrambling
IEA
GO_REF:0000108
ACCEPT
Summary: Loss of plasma membrane lipid asymmetry through TMEM63B-mediated bidirectional phospholipid translocation.
Reason: Core biological process, the direct consequence of the scramblase activity. Demonstrated both by phosphatidylserine exposure in cells and by altered steady-state phosphatidylcholine and sphingomyelin distribution in TMEM63B-deficient cells.
GO:0031901 early endosome membrane
IEA
GO_REF:0000044
ACCEPT
Summary: Early endosome membrane localisation.
Reason: Experimentally observed secondary location.
GO:0034220 monoatomic ion transmembrane transport
IEA
GO_REF:0000108
MODIFY
Summary: Generic monoatomic ion transport, derived electronically from the mechanosensitive ion channel term.
Reason: Correct but under-specific: TMEM63B conducts cations (Ca2+, Na+, K+, Mg2+, Cs+), and the cation-level term is already annotated.
GO:0098655 monoatomic cation transmembrane transport
IEA
GO_REF:0000120
ACCEPT
Summary: Cation flux across the membrane, the direct consequence of channel opening under mechanical or osmotic stress.
Reason: Core biological process for the channel activity.
GO:0007605 sensory perception of sound
IEA
GO_REF:0000107
KEEP AS NON CORE
Summary: Hearing phenotype transferred from mouse Tmem63b, where knockout causes outer hair cell loss and deafness.
Reason: A real organismal role of the orthologue, but a tissue-specific downstream consequence of the channel activity rather than a process the protein itself carries out. No human evidence.
GO:0008381 mechanosensitive monoatomic ion channel activity
IEA
GO_REF:0000107
MODIFY
Summary: Mechanosensitive ion channel activity, correct in kind but under-specific.
Reason: TMEM63B is specifically a mechanosensitive cation channel, and the cation-level term is available and already carried by this gene from an IDA. Using the parent term loses the ion-class information that the recordings establish.
GO:0017128 phospholipid scramblase activity
IEA
GO_REF:0000107
ACCEPT
Summary: TMEM63B translocates phospholipids bidirectionally across the bilayer in response to changes in membrane curvature and thickness.
Reason: Core molecular function, and one of two core activities. Phospholipid scrambling by TMEM63B is supported by four independent groups using four different systems: an unbiased CRISPR scrambling screen in pro-B cells (PMID:39217145), purified protein plus open/closed cryo-EM structures resolving a lipid translocation pathway (PMID:39424995), a paralog/ortholog survey in a Tmem63b-null background (PMID:39716028), and reconstitution into giant unilamellar vesicles with coarse-grained MD (PMID:41617699). GOA already carries both this activity and mechanosensitive cation channel activity with experimental evidence, and the two are not the same measurement re-described: disease variants dissociate them (V44M and T481N confer constitutive scrambling with no obvious effect on mechanosensitive channel activity, PMID:42573579, PMID:40480214), and ion conduction and lipid translocation have distinct bottleneck residues in the same groove (PMID:41617699). TMEM16 proteins are the family precedent for one protein being both a channel and a scramblase. What remains open is whether wild-type scrambling operates in vivo at physiological force rather than only in reconstituted or over-expressing systems; that question is recorded in suggested_questions and does not warrant demotion from core.
GO:0036335 intestinal stem cell homeostasis
IEA
GO_REF:0000107
KEEP AS NON CORE
Summary: Intestinal stem cell homeostasis, transferred from mouse Tmem63b.
Reason: Tissue-specific organismal phenotype of the mouse orthologue, downstream of mechanotransduction; not a core activity of the human protein and not demonstrated in human.
GO:0042756 drinking behavior
IEA
GO_REF:0000107
KEEP AS NON CORE
Summary: Drinking behaviour, from mouse Tmem63b acting as an osmosensor in subfornical organ neurons.
Reason: A genuine and well-documented organismal role of the mouse orthologue, but a behavioural output several steps downstream of the channel activity.
GO:0120019 phosphatidylcholine transfer activity
IEA
GO_REF:0000107
MODIFY
Summary: Phosphatidylcholine handling annotated with an intermembrane lipid-transfer term rather than a scramblase term.
Reason: Wrong mechanism class. GO:0120019 describes an intermembrane lipid-transfer protein: it 'removes phosphatidylcholine from a membrane or a monolayer lipid particle, transports it through the aqueous phase while protected in a hydrophobic pocket, and brings it to an acceptor membrane or lipid particle'. That is the STARD/CERT mode of action - a soluble carrier shuttling a lipid through the aqueous phase between two membranes. TMEM63B is an 11-transmembrane integral membrane protein that flips lipids between the two leaflets of a single bilayer through a membrane-spanning groove resolved by cryo-EM (PMID:39424995). The underlying observation is real - TMEM63B loss alters phosphatidylcholine and sphingomyelin distribution in the plasma membrane - but it was mapped to a transfer-protein term instead of a scramblase term. Replace with phospholipid scramblase activity, which the gene already carries with IDA and IMP evidence. GO currently has no lipid-species-specific scramblase child; a species-specific term is suggested in proposed_new_terms.
Proposed replacements: phospholipid scramblase activity
GO:0140338 sphingomyelin transfer activity
IEA
GO_REF:0000107
MODIFY
Summary: Sphingomyelin handling annotated with an intermembrane lipid-transfer term rather than a scramblase term.
Reason: Wrong mechanism class. GO:0140338 describes an intermembrane lipid-transfer protein: it 'removes a sphingomyelin from the outer leaflet of a donor membrane, transports it through the aqueous phase while protected in a hydrophobic pocket, and brings it to the outer leaflet of an acceptor membrane'. That is the STARD/CERT mode of action - a soluble carrier shuttling a lipid through the aqueous phase between two membranes. TMEM63B is an 11-transmembrane integral membrane protein that flips lipids between the two leaflets of a single bilayer through a membrane-spanning groove resolved by cryo-EM (PMID:39424995). The underlying observation is real - TMEM63B loss alters phosphatidylcholine and sphingomyelin distribution in the plasma membrane - but it was mapped to a transfer-protein term instead of a scramblase term. Replace with phospholipid scramblase activity, which the gene already carries with IDA and IMP evidence. GO currently has no lipid-species-specific scramblase child; a species-specific term is suggested in proposed_new_terms.
Proposed replacements: phospholipid scramblase activity
GO:0160069 surfactant secretion
IEA
GO_REF:0000107
KEEP AS NON CORE
Summary: Surfactant secretion from alveolar type 2 cells, downstream of stretch-activated TMEM63A/B currents.
Reason: A well-documented in vivo role - Tmem63a/b double knockout mice die of atelectasis and respiratory failure - but a tissue-specific secretory process downstream of the channel activity rather than something the protein does directly.
GO:1990760 osmolarity-sensing monoatomic cation channel activity
IEA
GO_REF:0000107
ACCEPT
Summary: Osmolarity-gated cation channel activity; the same conductance as the mechanosensitive channel activity, described by its osmotic gating stimulus.
Reason: Core. Hypo-osmotic activation of TMEM63B currents is directly demonstrated, and the osmotic and mechanical descriptions are two views of one force-from-lipids gating mechanism rather than two activities. Retained alongside GO:0140135.
GO:0005886 plasma membrane
IDA
GO_REF:0000052
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0015629 actin cytoskeleton
IDA
GO_REF:0000052
KEEP AS NON CORE
Summary: Colocalisation with cortical F-actin.
Reason: A reproducible association (reported both by HPA imaging and in the original HEK293T characterisation) rather than a compartment in which the channel acts; likely reflects cortical cytoskeleton coupling relevant to force transmission. Retained as non-core.
Supporting Evidence:
PMID:31243992
In HEK293T cells, TMEM63B localizes to the plasma membrane and is associated with F-actin.
GO:0005789 endoplasmic reticulum membrane
ISS
GO_REF:0000024
KEEP AS NON CORE
Summary: Endoplasmic reticulum membrane pool, supported by orthology to mouse Tmem63b.
Reason: Plausible and supported experimentally in mouse, but for a plasma-membrane/lysosomal channel-scramblase the ER pool most likely reflects the biosynthetic route rather than a site of action. Retained as a secondary location.
GO:0036335 intestinal stem cell homeostasis
ISS
GO_REF:0000024
KEEP AS NON CORE
Summary: Intestinal stem cell homeostasis, transferred from mouse Tmem63b.
Reason: Tissue-specific organismal phenotype of the mouse orthologue, downstream of mechanotransduction; not a core activity of the human protein and not demonstrated in human.
GO:0042756 drinking behavior
ISS
GO_REF:0000024
KEEP AS NON CORE
Summary: Drinking behaviour, from mouse Tmem63b acting as an osmosensor in subfornical organ neurons.
Reason: A genuine and well-documented organismal role of the mouse orthologue, but a behavioural output several steps downstream of the channel activity.
GO:1990760 osmolarity-sensing monoatomic cation channel activity
ISS
GO_REF:0000024
ACCEPT
Summary: Osmolarity-gated cation channel activity; the same conductance as the mechanosensitive channel activity, described by its osmotic gating stimulus.
Reason: Core. Hypo-osmotic activation of TMEM63B currents is directly demonstrated, and the osmotic and mechanical descriptions are two views of one force-from-lipids gating mechanism rather than two activities. Retained alongside GO:0140135.
GO:0005886 plasma membrane
IDA
PMID:39217145
Phospholipid scrambling induced by an ion channel/metabolite...
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0017128 phospholipid scramblase activity
IMP
PMID:39217145
Phospholipid scrambling induced by an ion channel/metabolite...
ACCEPT
Summary: TMEM63B translocates phospholipids bidirectionally across the bilayer in response to changes in membrane curvature and thickness.
Reason: Core molecular function, and one of two core activities. Phospholipid scrambling by TMEM63B is supported by four independent groups using four different systems: an unbiased CRISPR scrambling screen in pro-B cells (PMID:39217145), purified protein plus open/closed cryo-EM structures resolving a lipid translocation pathway (PMID:39424995), a paralog/ortholog survey in a Tmem63b-null background (PMID:39716028), and reconstitution into giant unilamellar vesicles with coarse-grained MD (PMID:41617699). GOA already carries both this activity and mechanosensitive cation channel activity with experimental evidence, and the two are not the same measurement re-described: disease variants dissociate them (V44M and T481N confer constitutive scrambling with no obvious effect on mechanosensitive channel activity, PMID:42573579, PMID:40480214), and ion conduction and lipid translocation have distinct bottleneck residues in the same groove (PMID:41617699). TMEM16 proteins are the family precedent for one protein being both a channel and a scramblase. What remains open is whether wild-type scrambling operates in vivo at physiological force rather than only in reconstituted or over-expressing systems; that question is recorded in suggested_questions and does not warrant demotion from core.
Supporting Evidence:
PMID:39217145
Ca2+ stimulation-mediated PLS is suppressed by deletion of Tmem63b, while human disease-related Tmem63b mutants induce constitutive PLS.
GO:0005227 calcium-activated cation channel activity
IDA
PMID:37421948
Stretch-activated ion channel TMEM63B associates with develo...
MODIFY
Summary: Ca2+-permeable, stretch/osmolarity-gated cation channel annotated with a term that asserts gating by Ca2+ binding.
Reason: The essence is right - TMEM63B is a Ca2+-permeable cation channel - but the term names the wrong gating stimulus. GO:0005227 is defined as a channel that opens 'when a calcium cation has been bound by the channel complex or one of its constituent parts', and every experimental user of the term in GOA is a genuinely Ca2+-gated channel (KCNN4, SLO1/SLO2, TRPM4, TMEM16A). TMEM63B is gated by membrane stretch and hypo-osmolarity, with Ca2+ as a permeant ion. The provenance of the slip is visible: the mouse IDA that seeds the human ISS/IEA rows comes from a paper titled 'Overexpression of Osmosensitive Ca(2+)-Permeable Channel TMEM63B' (PMID:31243992), and the InterPro family behind the IEA is named 'Calcium permeable stress-gated cation channel 1-like' (IPR045122) yet is mapped to the calcium-activated term. Replace with the mechanosensitive cation channel term plus a plain calcium channel term, which together capture what was actually measured without asserting Ca2+ gating.
Supporting Evidence:
PMID:37421948
demonstrated inward leak cation currents across the mutated channel even in isotonic conditions, while the response to hypo-osmotic challenge was impaired, as were the Ca2+ transients generated under hypo-osmotic stimulation
GO:0005765 lysosomal membrane
IDA
PMID:39716028
Membrane structure-responsive lipid scramblase activity of t...
ACCEPT
Summary: TMEM63B also resides in the lysosomal membrane.
Reason: Experimentally supported secondary location; the scramblase work localises TMEM63B to both the plasma membrane and lysosomes.
Supporting Evidence:
PMID:39716028
We expressed human TMEM63 paralogs, TMEM63B orthologs, and plant OSCA1.1 in Tmem63b-deficient mouse pro-B cells and found that vertebrate TMEM63B orthologs exhibit scramblase activity at the PM.
GO:0005886 plasma membrane
IDA
PMID:37421948
Stretch-activated ion channel TMEM63B associates with develo...
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0005886 plasma membrane
IDA
PMID:39716028
Membrane structure-responsive lipid scramblase activity of t...
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
Supporting Evidence:
PMID:39716028
We expressed human TMEM63 paralogs, TMEM63B orthologs, and plant OSCA1.1 in Tmem63b-deficient mouse pro-B cells and found that vertebrate TMEM63B orthologs exhibit scramblase activity at the PM.
GO:0017128 phospholipid scramblase activity
IMP
PMID:39424995
Membrane structure-responsive lipid scrambling by TMEM63B to...
ACCEPT
Summary: TMEM63B translocates phospholipids bidirectionally across the bilayer in response to changes in membrane curvature and thickness.
Reason: Core molecular function, and one of two core activities. Phospholipid scrambling by TMEM63B is supported by four independent groups using four different systems: an unbiased CRISPR scrambling screen in pro-B cells (PMID:39217145), purified protein plus open/closed cryo-EM structures resolving a lipid translocation pathway (PMID:39424995), a paralog/ortholog survey in a Tmem63b-null background (PMID:39716028), and reconstitution into giant unilamellar vesicles with coarse-grained MD (PMID:41617699). GOA already carries both this activity and mechanosensitive cation channel activity with experimental evidence, and the two are not the same measurement re-described: disease variants dissociate them (V44M and T481N confer constitutive scrambling with no obvious effect on mechanosensitive channel activity, PMID:42573579, PMID:40480214), and ion conduction and lipid translocation have distinct bottleneck residues in the same groove (PMID:41617699). TMEM16 proteins are the family precedent for one protein being both a channel and a scramblase. What remains open is whether wild-type scrambling operates in vivo at physiological force rather than only in reconstituted or over-expressing systems; that question is recorded in suggested_questions and does not warrant demotion from core.
Supporting Evidence:
PMID:39424995
we show that transmembrane protein 63B (TMEM63B) functions as a membrane structure-responsive lipid scramblase localized at the PM and lysosomes, activating bidirectional lipid translocation upon changes in membrane curvature and thickness.
GO:0017128 phospholipid scramblase activity
IDA
PMID:39716028
Membrane structure-responsive lipid scramblase activity of t...
ACCEPT
Summary: TMEM63B translocates phospholipids bidirectionally across the bilayer in response to changes in membrane curvature and thickness.
Reason: Core molecular function, and one of two core activities. Phospholipid scrambling by TMEM63B is supported by four independent groups using four different systems: an unbiased CRISPR scrambling screen in pro-B cells (PMID:39217145), purified protein plus open/closed cryo-EM structures resolving a lipid translocation pathway (PMID:39424995), a paralog/ortholog survey in a Tmem63b-null background (PMID:39716028), and reconstitution into giant unilamellar vesicles with coarse-grained MD (PMID:41617699). GOA already carries both this activity and mechanosensitive cation channel activity with experimental evidence, and the two are not the same measurement re-described: disease variants dissociate them (V44M and T481N confer constitutive scrambling with no obvious effect on mechanosensitive channel activity, PMID:42573579, PMID:40480214), and ion conduction and lipid translocation have distinct bottleneck residues in the same groove (PMID:41617699). TMEM16 proteins are the family precedent for one protein being both a channel and a scramblase. What remains open is whether wild-type scrambling operates in vivo at physiological force rather than only in reconstituted or over-expressing systems; that question is recorded in suggested_questions and does not warrant demotion from core.
Supporting Evidence:
PMID:39716028
We expressed human TMEM63 paralogs, TMEM63B orthologs, and plant OSCA1.1 in Tmem63b-deficient mouse pro-B cells and found that vertebrate TMEM63B orthologs exhibit scramblase activity at the PM.
GO:0120019 phosphatidylcholine transfer activity
ISS
GO_REF:0000024
MODIFY
Summary: Phosphatidylcholine handling annotated with an intermembrane lipid-transfer term rather than a scramblase term.
Reason: Wrong mechanism class. GO:0120019 describes an intermembrane lipid-transfer protein: it 'removes phosphatidylcholine from a membrane or a monolayer lipid particle, transports it through the aqueous phase while protected in a hydrophobic pocket, and brings it to an acceptor membrane or lipid particle'. That is the STARD/CERT mode of action - a soluble carrier shuttling a lipid through the aqueous phase between two membranes. TMEM63B is an 11-transmembrane integral membrane protein that flips lipids between the two leaflets of a single bilayer through a membrane-spanning groove resolved by cryo-EM (PMID:39424995). The underlying observation is real - TMEM63B loss alters phosphatidylcholine and sphingomyelin distribution in the plasma membrane - but it was mapped to a transfer-protein term instead of a scramblase term. Replace with phospholipid scramblase activity, which the gene already carries with IDA and IMP evidence. GO currently has no lipid-species-specific scramblase child; a species-specific term is suggested in proposed_new_terms.
Proposed replacements: phospholipid scramblase activity
Supporting Evidence:
PMID:39424995
TMEM63B deficiency alters phosphatidylcholine and sphingomyelin distributions in the PM.
GO:0140338 sphingomyelin transfer activity
ISS
GO_REF:0000024
MODIFY
Summary: Sphingomyelin handling annotated with an intermembrane lipid-transfer term rather than a scramblase term.
Reason: Wrong mechanism class. GO:0140338 describes an intermembrane lipid-transfer protein: it 'removes a sphingomyelin from the outer leaflet of a donor membrane, transports it through the aqueous phase while protected in a hydrophobic pocket, and brings it to the outer leaflet of an acceptor membrane'. That is the STARD/CERT mode of action - a soluble carrier shuttling a lipid through the aqueous phase between two membranes. TMEM63B is an 11-transmembrane integral membrane protein that flips lipids between the two leaflets of a single bilayer through a membrane-spanning groove resolved by cryo-EM (PMID:39424995). The underlying observation is real - TMEM63B loss alters phosphatidylcholine and sphingomyelin distribution in the plasma membrane - but it was mapped to a transfer-protein term instead of a scramblase term. Replace with phospholipid scramblase activity, which the gene already carries with IDA and IMP evidence. GO currently has no lipid-species-specific scramblase child; a species-specific term is suggested in proposed_new_terms.
Proposed replacements: phospholipid scramblase activity
Supporting Evidence:
PMID:39424995
TMEM63B deficiency alters phosphatidylcholine and sphingomyelin distributions in the PM.
GO:0160069 surfactant secretion
ISS
GO_REF:0000024
KEEP AS NON CORE
Summary: Surfactant secretion from alveolar type 2 cells, downstream of stretch-activated TMEM63A/B currents.
Reason: A well-documented in vivo role - Tmem63a/b double knockout mice die of atelectasis and respiratory failure - but a tissue-specific secretory process downstream of the channel activity rather than something the protein does directly.
Supporting Evidence:
PMID:38127458
we show that loss of the mechanosensitive channels TMEM63A and TMEM63B (TMEM63A/B) resulted in atelectasis and respiratory failure in mice due to a deficit of surfactant secretion
GO:0005886 plasma membrane
IDA
PMID:38127458
Mechanosensitive channels TMEM63A and TMEM63B mediate lung i...
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0031901 early endosome membrane
IDA
PMID:38127458
Mechanosensitive channels TMEM63A and TMEM63B mediate lung i...
ACCEPT
Summary: Early endosome membrane localisation.
Reason: Experimentally observed secondary location.
GO:0097233 alveolar lamellar body membrane
IDA
PMID:38127458
Mechanosensitive channels TMEM63A and TMEM63B mediate lung i...
ACCEPT
Summary: Limiting membrane of the alveolar lamellar body, a lysosome-related organelle.
Reason: Specific and experimentally demonstrated location in alveolar type 2 cells, and the site at which TMEM63B triggers lamellar body exocytosis.
Supporting Evidence:
PMID:38127458
TMEM63A/B were predominantly localized at the limiting membrane of the lamellar body (LB), a lysosome-related organelle that stores pulmonary surfactant and ATP in AT2 cells.
GO:0140135 mechanosensitive monoatomic cation channel activity
IDA
PMID:38127458
Mechanosensitive channels TMEM63A and TMEM63B mediate lung i...
ACCEPT
Summary: Direct recording of stretch-activated cation currents carried by TMEM63B.
Reason: Core molecular function. TMEM63B is a bona fide mechanically activated cation channel with small conductance and a high activation threshold, established by heterologous electrophysiology (PMID:30382938, PMID:37543036), by stretch-evoked cation currents in alveolar epithelial cells in vivo (PMID:38127458), and by the altered conductance of human disease variants (PMID:37421948). This activity coexists with, and is separable from, the phospholipid scramblase activity the gene also carries.
Supporting Evidence:
PMID:38127458
Activation of TMEM63A/B channels during cell stretch facilitated the release of surfactant and ATP from LBs fused with the plasma membrane.
GO:0005886 plasma membrane
IDA
PMID:37543036
TMEM63 proteins function as monomeric high-threshold mechano...
ACCEPT
Summary: TMEM63B is a multi-pass plasma membrane protein and both its channel currents and its scrambling are recorded at the plasma membrane.
Reason: Correct and well-supported primary location, reported independently by structural, electrophysiological, scramblase and HPA studies.
GO:0007605 sensory perception of sound
ISS
GO_REF:0000024
KEEP AS NON CORE
Summary: Hearing phenotype transferred from mouse Tmem63b, where knockout causes outer hair cell loss and deafness.
Reason: A real organismal role of the orthologue, but a tissue-specific downstream consequence of the channel activity rather than a process the protein itself carries out. No human evidence.
GO:0008381 mechanosensitive monoatomic ion channel activity
IDA
PMID:37543036
TMEM63 proteins function as monomeric high-threshold mechano...
MODIFY
Summary: Mechanosensitive ion channel activity, correct in kind but under-specific.
Reason: TMEM63B is specifically a mechanosensitive cation channel, and the cation-level term is available and already carried by this gene from an IDA. Using the parent term loses the ion-class information that the recordings establish.
Supporting Evidence:
PMID:37543036
Functional analyses demonstrated that TMEM63s are bona fide mechanosensitive ion channels, characterized by small conductance and high thresholds.
GO:0005227 calcium-activated cation channel activity
ISS
GO_REF:0000024
MODIFY
Summary: Ca2+-permeable, stretch/osmolarity-gated cation channel annotated with a term that asserts gating by Ca2+ binding.
Reason: The essence is right - TMEM63B is a Ca2+-permeable cation channel - but the term names the wrong gating stimulus. GO:0005227 is defined as a channel that opens 'when a calcium cation has been bound by the channel complex or one of its constituent parts', and every experimental user of the term in GOA is a genuinely Ca2+-gated channel (KCNN4, SLO1/SLO2, TRPM4, TMEM16A). TMEM63B is gated by membrane stretch and hypo-osmolarity, with Ca2+ as a permeant ion. The provenance of the slip is visible: the mouse IDA that seeds the human ISS/IEA rows comes from a paper titled 'Overexpression of Osmosensitive Ca(2+)-Permeable Channel TMEM63B' (PMID:31243992), and the InterPro family behind the IEA is named 'Calcium permeable stress-gated cation channel 1-like' (IPR045122) yet is mapped to the calcium-activated term. Replace with the mechanosensitive cation channel term plus a plain calcium channel term, which together capture what was actually measured without asserting Ca2+ gating.
Supporting Evidence:
PMID:31243992
This Ca2+-permeable channel specifically induces Ca2+ influx across the membrane in response to extracellular Ca2+ concentration and hyperosmolarity.
GO:0008381 mechanosensitive monoatomic ion channel activity
ISS
GO_REF:0000024
MODIFY
Summary: Mechanosensitive ion channel activity, correct in kind but under-specific.
Reason: TMEM63B is specifically a mechanosensitive cation channel, and the cation-level term is available and already carried by this gene from an IDA. Using the parent term loses the ion-class information that the recordings establish.

Core Functions

TMEM63B forms a monomeric, small-conductance, high-threshold cation channel that opens in response to membrane tension and to hypo-osmotic swelling. It is permeable to Ca2+, Na+, K+ and Mg2+, and gates by the force-from-lipids mechanism rather than by ligand binding, so the osmolarity-sensing description (GO:1990760) and the mechanosensitive description refer to one activity. In alveolar type 2 cells this conductance converts lung inflation into the Ca2+ signal that drives lamellar body exocytosis.

Supporting Evidence:
  • PMID:37543036
    Functional analyses demonstrated that TMEM63s are bona fide mechanosensitive ion channels, characterized by small conductance and high thresholds.
  • PMID:38127458
    Activation of TMEM63A/B channels during cell stretch facilitated the release of surfactant and ATP from LBs fused with the plasma membrane.
  • PMID:30382938
    Here, we show that various members of the OSCA and TMEM63 family of proteins from plants, flies, and mammals confer mechanosensitivity to naïve cells.

TMEM63B translocates phospholipids bidirectionally between the two leaflets of the bilayer through a membrane-spanning hydrophilic groove, without ATP, when membrane curvature and thickness change or when tension is applied. Scrambling dissipates plasma membrane lipid asymmetry, exposing phosphatidylserine and redistributing phosphatidylcholine and sphingomyelin. The lipid pathway is resolved in open and closed cryo-EM structures and lies in the same groove as the ion pore, but is controlled by different bottleneck residues, so the scramblase and channel activities are separable; pathogenic variants around the hydrophobic gate make the scrambling constitutive without enhancing ion conduction. It remains to be shown that wild-type scrambling is deployed at physiological force in vivo.

Supporting Evidence:
  • PMID:39424995
    we show that transmembrane protein 63B (TMEM63B) functions as a membrane structure-responsive lipid scramblase localized at the PM and lysosomes, activating bidirectional lipid translocation upon changes in membrane curvature and thickness.
  • PMID:39424995
    We determined the cryo-electron microscopy structures of TMEM63B in its open and closed conformations, uncovering a lipid translocation pathway formed in response to changes in the membrane environment.
  • PMID:39217145
    Ca2+ stimulation-mediated PLS is suppressed by deletion of Tmem63b, while human disease-related Tmem63b mutants induce constitutive PLS.
  • PMID:41617699
    We show that lipid scrambling in TMEM63 proteins can be activated by mechanical forces in the membrane, making these mechanically activated lipid scramblases.

References

Loading supporting content…

Download this section (compressed HTML)

Suggested Questions for Experts

Q: Is the phospholipid scramblase activity of wild-type TMEM63B deployed physiologically in vivo, or is it only detectable in reconstituted vesicles and over-expressing cells? Zheng et al. state this explicitly as an open question.

Suggested experts: Jeffrey R Holt, Katsumori Segawa, Huanghe Yang, Jun Suzuki

Q: Are the channel and scramblase activities of TMEM63B ever engaged simultaneously in a native cell, or are they alternative outcomes of the same gating transition selected by the magnitude and kind of mechanical stimulus? The I475del variant, which potentiates scrambling under hypotonic stress, suggests the two can be coupled.

Suggested experts: Huanghe Yang, Charles D Cox

Q: Does the Tmem63b/Slc19a2 heterodimer requirement reported for Ca2+-stimulated scrambling in pro-B cells apply to the purified-protein and GUV systems, where TMEM63B scrambles on its own? If not, what does SLC19A2 contribute?

Suggested experts: Jun Suzuki, Katsumori Segawa

Q: Which of the two activities accounts for which part of the TMEM63B channelopathy phenotype - the neurological features, or the macrocytic anaemia and haemolysis seen in a subset of individuals?

Suggested experts: Renzo Guerrini, Annalisa Vetro, Huanghe Yang

Suggested Experiments

Experiment: Apply calibrated membrane tension to cells expressing endogenous TMEM63B (for example alveolar type 2 cells or the subfornical organ neurons in which it is enriched) using high-speed pressure clamp or cyclic uniaxial stretch, while simultaneously recording annexin V binding and single-channel currents. Compare TMEM63B-null cells and cells rescued with scramble-dead but conduction-competent groove mutants.

Hypothesis: Wild-type TMEM63B scrambles phospholipids in native cells only above a tension threshold that is reached during physiological mechanical stimulation.

Type: combined electrophysiology and lipid-exposure imaging under controlled mechanical load

Experiment: Generate knock-in mice carrying groove mutations that selectively abolish lipid translocation while preserving stretch-activated current, and vice versa, using the separable bottleneck residues identified by MD. Phenotype for surfactant secretion, hearing, thirst behaviour and erythrocyte indices.

Hypothesis: Ion conduction and lipid scrambling can be separated in vivo by bottleneck-residue mutations, and each contributes to a distinguishable part of the disease phenotype.

Type: mouse knock-in structure-function genetics

Experiment: Reconstitute purified TMEM63B into proteoliposomes of defined composition and measure leaflet-specific translocation rates for NBD-PC, NBD-SM and NBD-PS by dithionite quenching, with and without an acceptor vesicle population. An acceptor-independent rate would formally exclude the GO:0120019/GO:0140338 transfer-protein mechanism that the current annotations assert.

Hypothesis: TMEM63B scrambles phosphatidylcholine and sphingomyelin with kinetics distinguishable from an intermembrane lipid-transfer protein.

Type: proteoliposome scramblase assay

📚 Additional Documentation

Notes

(TMEM63B-notes.md)

Loading supporting content…

Download this section (compressed HTML)

📄 View Raw YAML

Loading supporting content…

Download this section (compressed HTML)