CASTOR (Lotus japonicus, Q5H8A6) – curation notes

Sources

• Deep research: CASTOR-deep-research-falcon.md (present).
• Cached: PMID:19106374 and PMID:31420535 full text; PMID:15616514, 16903357, 22706284, 27230377, 39140702 abstract only.
• DOIs resolved to PMIDs with the PubMed MCP (31420535, 22706284, 27230377, 39140702).

Identity and localization

• Gene identified with its twin POLLUX as a common symbiosis gene PMID:15616514.
• The plastid localization in that paper came from overexpressed fusions. Native protein is at the nuclear envelope PMID:19106374. Inner versus outer membrane could not be resolved PMID:19106374.
• CASTOR and POLLUX form separate homocomplexes PMID:19106374.

Channel activity: the K+ vs Ca2+ debate

• Full-length CASTOR in bilayers: 260 pS, weak cation selectivity, K+ preferred PMID:19106374. The authors infer K+ is the main ion in vivo and propose a counter-ion or membrane-potential role PMID:19106374.
• Kim et al. 2019: a truncated, plasma-membrane-targeted pore construct (CASTOR-2TM) is Ca2+-selective and Ca2+-activated PMID:31420535. The RCK gating-ring structure and the nodulation-rescue failure of Ca2+-site mutants are solid.
• CNGC15 is the nuclear-envelope Ca2+ channel and complexes with the K+-permeable DMI1 PMID:27230377.
• 2024 review: the question is still open PMID:39140702.
• Decision: MF annotated at the level of monoatomic cation channel activity (GO:0005261). K+ transport is kept because full-length bilayer data support it. Calcium channel activity is not asserted.

Genetics

• castor mutants lack Ca2+ spiking but keep the initial Ca2+ flux PMID:16903357.
• castor-2 (pore-region A264T) is defective in both symbioses PMID:19106374.
• Channel-dead Ca2+-site mutants fail to rescue nodulation PMID:31420535.
• DMI1 alone rescues Lotus castor/pollux mutants, while Lotus needs both channels PMID:22706284.

NEW-term checks

• Nodulation (GO:0009877) and AM association (GO:0036377): CASTOR's channel conductance is a mechanistic step in generating nuclear Ca2+ spikes, so it participates and is not merely required. Comparator check (QuickGO, taxa 3880/34305): CNGC15A/CNGC15B (IMP), CCaMK (IMP) and CYCLOPS (IMP/IBA) carry these terms. These are proteins in the same role (nuclear Ca2+ spiking machinery or its decoders), so the term absence for CASTOR is a gap, not a convention.
• GOA has only 3 IEA rows for CASTOR (no MF, no experimental rows), so these additions are not overruling any curator.

Project relevance (symbiosis vs defense)

• CASTOR is purely a symbiosis-pathway component. There is no evidence for a role in fungal or bacterial defense, and no defense terms are proposed.
• Medicago DMI1 (POLLUX ortholog) is reviewed separately; DMI1-specific experiments were not transferred to CASTOR.