SAGA2 (Chlamydomonas reinhardtii) curation notes

UniProt A0A2K3DEH1 (TrEMBL, "CBM20 domain-containing protein"); locus Cre09.g394621
(CHLRE_09g394621v5); 1,816 aa, 189.5 kDa predicted.

Second UniProt entry. A0A2K3DEF8 (TrEMBL, 1,748 aa, "CBM20 domain-containing
protein") carries the same ORF name CHLRE_09g394621v5 and EMBL CM008970 (checked by
UniProt REST query on 2026-10-04). The two entries are alternative gene-model products of
one gene. This review uses A0A2K3DEH1, the entry that carries the GOA annotations. Which
model is correct is unresolved.

Deep research

Deep research was not run. The falcon provider returns HTTP 402 and perplexity is
unavailable in this environment, so the wrapper was skipped on instruction. No
-deep-research-*.md file exists for this gene. The review rests on the cached primary
literature below.

Sequence features (UniProt record)

• Two consecutive CBM20 starch-binding domains, residues 96-206 and 221-343 (PROSITE PS51166; Pfam PF00686; InterPro IPR002044).
• Coiled-coil regions 714-1244; extensive disordered, low-complexity regions (Pro-, Gln-rich) including the C-terminal 1403-1816.
• No predicted transmembrane segment.
• PANTHER PTHR15048:SF0 (STARCH-BINDING DOMAIN-CONTAINING PROTEIN 1), the STBD1 family. The two IBA rows come from this family.
• Four Rubisco-binding motifs PMID:38241434.

Literature summary

Meyer et al. 2020 (PMID:33177094): naming, motif, localization

• Named here PMID:33177094; it co-precipitated with the anti-SAGA1 antibody because it carries the same Rubisco-binding motif.
• Rubisco bound SAGA2 internal motif peptides on arrays PMID:33177094.
• SAGA2-Venus at the matrix-starch interface PMID:33177094.

Atkinson/Chan et al. 2024 (PMID:38241434): Arabidopsis proto-pyrenoid

• SAGA2 transit peptide targets chloroplast stroma PMID:38241434.
• Y2H with Rubisco LSU/SSU and EPYC1 helices PMID:38241434.
• Recruits starch to the condensate, and binds starch outside it as well PMID:38241434.
• Note: this paper calls the CBM20 C-terminal; UniProt puts both CBM20s at the N terminus.

Hennacy et al. 2024 (PMID:39548241)

• SAGA2 excluded from tubule biogenesis PMID:39548241.

Crans et al. 2026 (PNAS PMID:42090253): primary SAGA2 genetics and biochemistry

• Direct starch binding by the purified two-CBM20 construct PMID:42090253, competed by beta-cyclodextrin and maltoheptaose.
• SAGA2 is depleted from tubule entry sites, where SAGA1 is enriched PMID:42090253.
• saga2 grows normally, has a single pyrenoid and normal tubules PMID:42090253.
• saga2 has reduced starch coverage at 3 h of low CO2 (33% vs WT), rescued by SAGA2-Venus, normal by 21 h PMID:42090253.
• Double mutant has no sheath; starch in the stroma; either gene rescues PMID:42090253.
• Starch granule initiation is proposed as a model only; not annotated.

Decisions

1. IBA vacuolar transport and membrane (PTN000386078, STBD1 family): REMOVE. The
   node is seeded by mammalian STBD1 (ER-anchored, glycophagy). SAGA2 shares only the CBM20,
   has no membrane anchor, is chloroplast-targeted and acts at the matrix-starch interface,
   away from membranes. Target-specific evidence of divergence, recorded in propagation_review.
2. IEA carbohydrate binding, starch binding: ACCEPT. Confirmed by the direct binding
   assay (domain-level). Consistent with SAGA1.
3. NEW pyrenoid (GO:1990732) IDA from SAGA2-Venus localization.
4. No process term. GO has no starch sheath assembly/localization term. GO:0033037
   polysaccharide localization exists but QuickGO shows only 2 annotations (Chst11,
   glycosaminoglycan), so it is not an established home for starch tethers. Raised as a
   suggested question, consistent with the SAGA1 review. Tubule terms (GO:0160223,
   GO:0010027) are explicitly not applicable to SAGA2.

Module (pyrenoid_ccm, starch_sheath_barrier)

SAGA2 is not a participant yet. Proposed annoton, mirroring saga1_matrix_starch_linker:
participant UniProtKB:A0A2K3DEH1; function GO:2001070 starch binding; location GO:1990732
pyrenoid; no process.