DMI1 (Medicago truncatula, Q6RHR6) – curation notes

Sources

UniProt Q6RHR6 (Swiss-Prot), GOA (7 rows), falcon deep research
(DMI1-deep-research-falcon.md), cached PMIDs below.
Full text available: PMID:35972963 (Liu 2022), PMID:39814887 (Cook 2025),
PMID:31420535 (Kim 2019, Lotus CASTOR). Abstract only: PMID:14963334, 17173544,
17631529, 21825141, 22706284, 27230377, 39140702.
PMIDs for Capoen 2011, Liu 2022 and Cook 2025 were resolved from DOI via Europe PMC
(PubMed MCP was rate-limited).

POLLUX-type member of the CASTOR/POLLUX (TC 1.A.1.23) family; 4 predicted TM helices,
two cytosolic RCK domains (RCK1 348-489, RCK2 608-757); crystal structure of the
C-terminal soluble region (PDB 7VM8). PMID:14963334
RCK = MthK-like Ca2+-regulated gating ring PMID:17631529.
NAD(P)-binding Rossmann superfamily hit reflects RCK fold, not enzymatic activity.

Nuclear envelope, functional GFP fusion PMID:17173544.
Preferentially inner nuclear membrane PMID:21825141.
Charpentier 2016 (GOA EXP source) also places the DMI1-CNGC15 complex at the nuclear
envelope.

Bilayer: ~64 pS in symmetrical KCl, long open state; A-for-S filter substitution made
DMI1 "solo sufficient" vs Lotus CASTOR+POLLUX pair [PMID:22706284; deep research].
Charpentier 2016 describes DMI1 as "the postassium-permeable channel, which modulates
nuclear Ca(2+) release" PMID:27230377.
Cook 2025: the native filter (ADAGN) is non-selective cationic; replacing it with a
strict K+ filter (TVGYG) fully complements dmi1-1 PMID:39814887. So K+ conduction is sufficient for function.
Dispute: Kim 2019 reports Lotus CASTOR as a Ca2+-regulated Ca2+ channel
PMID:31420535. Homolog evidence only; Cook 2025 TVGYG data argue that DMI1 Ca2+
permeation is not required.

dmi1 mutants lack both nodulation and AM symbiosis; no Nod-factor calcium spiking
[PMID:14963334; UniProt disruption phenotype].
DMI1 forms a complex with CNGC15a/b/c (the Ca2+-permeable release channels)
PMID:27230377; GOA IPI x3 (protein binding) come from this.
Gain-of-function DMI1 (S760N, spd1) gives spontaneous oscillations and nodules,
CNGC15-dependent PMID:35972963.
Pacemaker: DMI1 is required for CNGC15 activation and sets the oscillation frequency
PMID:39814887.
Peiter 2007: DMI1 modulates Ca2+ release from stores in yeast; "DMI1 is not directly
responsible for Nod factor-induced calcium changes, but does have the capacity to
regulate calcium channels".

protein binding (IPI x3, CNGC15a/b/c) -> MODIFY to ion channel regulator activity
(GO:0099106; "modulates the activity of a channel via direct interaction with it").
Justified by direct interaction (C-terminus of DMI1 with N-terminus of CNGC15) plus
Cook 2025 showing DMI1 C-terminal Ca2+-binding mobility modulates CNGC15 gating.
monoatomic ion transport (IEA) -> MODIFY to monoatomic cation transmembrane transport.
potassium ion transport (IEA) -> ACCEPT (supported by bilayer, Charpentier 2016, TVGYG).
nuclear membrane (EXP, IEA) -> ACCEPT.
NEW: monoatomic cation channel activity (GO:0005261, IDA PMID:22706284), kept generic to match Lotus CASTOR review.
NEW: nodulation (GO:0009877) and arbuscular mycorrhizal association (GO:0036377).
Comparator check: CNGC15A and CNGC15B (the partner channels in the same oscillator,
same paper) carry IMP nodulation and IMP AM association; CCAMK and CYCLOPS carry
nodulation. Participation: DMI1 is a working component of the nuclear Ca2+ oscillator
(channel + pacemaker), not a substrate. Absence for DMI1 looks like a gap rather than a
convention.
Did not add nuclear inner membrane as NEW (descendant of existing nuclear membrane);
described in text instead.

DMI1 has no defense/immunity annotations; all roles are symbiotic signalling. Cook 2025
/ Jacott 2024 mention CNGC15/DMI1 roles beyond symbiosis (nitrate signalling) but not
defense; not annotated.