LHP1 (Q946J8, At5g17690; TFL2) curation notes

2026-10-06 initial review

• Deep research: just deep-research-falcon ARATH LHP1 failed (HTTP 402 from provider); review built from cached publications and UniProt.
• Domain architecture: N-terminal chromodomain (108-167) and C-terminal chromo shadow domain (378-442) [file:ARATH/LHP1/LHP1-uniprot.txt "Chromo 1"]. Single plant HP1 homolog.
• Core MF: H3K27me3 reader. PMID:17542647; PMID:17676062
• Complex: plant PRC1-like with RING1A/B PMID:19097900; bridged to PRC2 via MSI1 PMID:23778966. Added NEW part_of PRC1 complex (GO:0035102).
• Vernalization: required for maintenance (not establishment) of FLC silencing after cold PMID:16682972; PMID:16549797.
• Localization: mostly euchromatic foci, not chromocenters, for a complementing fusion PMID:16244868. Heterochromatin rows kept as non-core.

Mis-attributed GOA rows (PMID:23203051)

The RIBA paper (full text cached) studies RIBA1 (At5g64300), RIBA2 (At2g22450), RIBA3 (At5g59750); it never mentions LHP1/TFL2 (At5g17690). GTP cyclohydrolase II activity (IDA), NOT DHBPS activity (IDA) and chloroplast (IDA) on LHP1 are gene-attribution errors -> REMOVE. These should be reported to TAIR.

Other decisions

• Bare protein-binding rows: REMOVE per repo policy (no informative replacement), noting the interactions are not doubted. Self-interaction -> MODIFY to protein homodimerization activity PMID:21304947.
• Sequence-specific DNA binding (IDA PMID:27495811 and the derived IBA): motif enrichment in ChIP regions is not direct sequence-specific binding; LHP1 lacks a DNA-binding domain -> MARK_AS_OVER_ANNOTATED.
• DamID "DNA binding" -> MODIFY to chromatin binding.
• Module vernalization_flc_silencing lists LHP1 without MF; GO:0061628 (histone H3K27me3 reader activity) is now the core MF in this review.