SnRK2.3 (SRK2I, At5g66880, UniProt Q39193) curation notes

2026-10-19 Session 1: new review

Setup and sources

• Folder created with just fetch-gene ... --alias SnRK2.3; UniProt names the gene SRK2I.
  The review uses gene_symbol SnRK2.3 to match the folder, following the
  genes/ARATH/OST1 precedent (UniProt SRK2E, gene_symbol OST1).
• No deep-research provider is available in this environment; this journal records my own
  literature reading of the cached publications (just fetch-gene-pmids ARATH SnRK2.3)
  plus PubMed searches (esearch for "SRK2I", "SnRK2.3 phosphorylation", the 2009 triple
  mutant paper). Additional PMIDs cached: 19420218, 19541597, 22160701, 30962512,
  29659577, 28787436, 26943172.
• Abstract-only in cache: 19805022, 19880399, 10634783, 15292193, 21166475, 19420218,
  19541597, 22160701, 30962512. Full text: 17307925, 19924127, 23290725, 19874541,
  18941053 (introduction and discussion only), 21220313 (discussion), 28438792,
  28887381, 29659577, 28787436, 26943172.

Identity and family

• 361 aa subclass III SnRK2 (SNF1-related protein kinase 2) with a protein kinase domain
  (22-278), ATP-binding loop 28-36, Lys51 and catalytic Asp141 (UniProt features, PROSITE
  PRU00159). PANTHER PTHR24343:SF566. Crystal structure of full-length SnRK2.3 (PDB 3UC3)
  PMID:22160701.
• Closest paralogs: SnRK2.2/SRK2D (At3g50500) and SnRK2.6/OST1/SRK2E (At4g33950).

Activity and activation

• ABA- and osmotic-stress activated kinase [PMID:15292193 "Nine of the 10 SnRK2 were
  activated by hyperosmolarity induced by mannitol, as well as NaCl, indicating an
  important role of the SnRK2 family in osmotic signaling"; "abscisic acid only activated
  five of the nine SnRK2"].
• In-gel kinase assays with ABF2, ABI5 and ABF1 fragments: the 42-kD ABA-induced activity
  is from SnRK2.2/2.3 PMID:17307925.
• Two-step autoactivation: PP2C release, then activation-loop autophosphorylation
  [PMID:22160701 "This stabilization enables SnRK2s to gain full activity by activation
  loop autophosphorylation"; "Autophosphorylation is more efficient in SnRK2.6, which has
  higher stability than SnRK2.3"]. UniProt: "Autophosphorylated in vitro" (ECO:0000269,
  PubMed:22160701).
• Substrates beyond ABFs: AKS bHLH factors PMID:28438792;
  HAT1 HD-ZIP PMID:29659577; FREE1 PMID:30962512.

Negative regulators

• Clade A PP2Cs (ABI1, ABI2, HAB1, AHG1) bind and dephosphorylate it PMID:19805022; ABI1 in planta PMID:19874541. Y2H with ABI1/ABI2/HAB1
  PMID:19924127. HDX of SnRK2.3-HAB1 (PMID:23290725).
• TOPP1/AtI-2 PMID:26943172.
• SCF(AtPP2-B11) targets SnRK2.3 for proteasomal degradation PMID:28787436.
• All of these regulate SnRK2.3; binding of a PP2C is the target side of the interaction,
  so bare protein binding rows are removed and I do not propose protein phosphatase
  binding as a SnRK2.3 molecular function (consistent with the OST1 review).

Localization

• Nucleus and cytoplasm (IDA, PMID:19880399) PMID:19880399; in seeds mainly nuclear PMID:19541597; BiFC with ABI1 in nucleus and cytosol PMID:19924127.
• NOT cytosol (RCA, PMID:21166475) is an inference from absence in a cell-suspension
  cytosolic proteome; contradicted by the BiFC and IDA data. Remove (as OST1).
• Plastid (HDA, PMID:28887381): membrane-fraction correlation profiling; the cached text
  does not mention SnRK2.3 and no other evidence places it in plastids. Over-annotation
  (same as the other ARATH reviews citing this dataset for plastid).

Physiology

• snrk2.2 snrk2.3 double mutant: ABA-insensitive germination, root growth, dormancy, Pro
  accumulation and ABRE gene expression PMID:17307925; only minor
  stomatal role PMID:17307925.
• snrk2.2/2.3/2.6 triple mutant nearly ABA-blind PMID:19420218;
  drought-sensitive PMID:19880399;
  seed development and dormancy (PMID:19541597).
• Osmotic stress (decuple mutant) PMID:21220313.
• GA/germination IGI from Piskurewicz 2008 (PMID:18941053, with SnRK2.2): cached text
  lacks the Results; abstract says "The endogenous ABI5 phosphorylation and inhibition of
  germination could be recapitulated by the addition of a SnRK2 protein kinase to the ABI5
  overexpression line." Kept the GA row as non-core (deferring to the curator).

TAS rows from PMID:10634783

• Li et al. 2000 cloned Vicia faba AAPK, an OST1-like SnRK2 PMID:10634783. The paper is
  not about SnRK2.3 itself, but the asserted claims (protein kinase activity; ABA
  signaling) are correct for SnRK2.3 and are independently supported by PMID:17307925.
  Accept; worth flagging to TAIR that an SnRK2.3-specific reference would be better.

Decisions summary

• MF: Ser/Thr protein kinase activity is core (IBA/IEA accept; IDA 'kinase activity' to
  MODIFY -> GO:0004674). ATP binding accept (as OST1). Protein binding: all REMOVE.
• BP: ABA-activated signaling pathway, response to ABA, response to water deprivation,
  response to osmotic stress accepted; salt non-core (as OST1); regulation of seed
  germination MODIFY -> negative regulation of seed germination (GO:0010187), since
  SnRK2.2/2.3 mediate ABA inhibition of germination.
• NEW: protein autophosphorylation (GO:0046777), PMID:22160701 + UniProt PTM; matches
  OST1 accepted IDA row.

2026-10-20: ported fixes from the parallel review branch

• Core function 1 now names FREE1 phosphorylation (curated UniProt line, PMID:30962512) and TOPP1/AtI-2 negative regulation PMID:26943172.
• The eight GO:0005515 rows with clade A PP2C partners changed from REMOVE to MODIFY to GO:0019903, as for SnRK2.2 and matching the OST1 GO:0019903 IPI row. The bHLH122 row stays REMOVE.
• GO:0009737 rows: reason explains why the term is kept alongside GO:0009738.
• The NOT located_in cytosol RCA row (PMID:21166475) stays REMOVE: it is contradicted by BiFC and IDA cytoplasm evidence. It is not recorded as a source defect, because it is a computational prediction rather than a curation error.

2026-10-25 — ABA module member audit

Assessed all 37 existing annotation decisions and core synthesis. The redundant seed/seedling ABA kinase role remains supported by cited primary studies. Description, core function and proposed autophosphorylation annotation are qualified with PMID:33911084: "transphosphorylation by RAFs is required for the reactivation of SnRK2.2/2.3/2.6." Its direct SnRK2.3 analog-sensitive assays distinguish RAF initiation from amplification; PMID:31896774 corroborates the upstream MAP3K requirement. These findings do not invalidate the recombinant-kinase/structure paper PMID:22160701, but limit the self-reactivation interpretation.

The plastid HDA call from PMID:28887381 is now UNDECIDED. Failure to find SnRK2.3 in article running text does not refute supplementary proteomics evidence; a nuclear/cytoplasmic function does not exclude a second pool. The gene-specific fractionation assignment remains to be checked. The negated cytosol RCA assertion remains rejected against independent cytosolic interaction evidence. PP2C-binding rationales no longer depend on deduplication.

Research used existing provider reports/manual notes and cached primary literature. All rows retain explicit decisions; unresolved findings are UNDECIDED rather than inferred from missing text. Record-section backfill and targeted validation were run for this session.