At3g08840

Summary: Electronic nucleotide-binding annotation from a UniRule. The protein has two ATP-grasp domains, and the nucleotide it binds is ATP, which this gene already carries as GO:0005524 ATP binding.

Reason: Correct but less specific than the evidence supports. The ATP-grasp domains of Ddl enzymes bind ATP, which is consumed when the two D-Ala molecules are joined; the specific term GO:0005524 ATP binding is already annotated to this gene.

Proposed replacements: ATP binding

Summary: ATP binding inferred from the two ATP-grasp domains (PROSITE PS50975, residues 223-452 and 672-925). Ddl enzymes use ATP to activate the first D-Ala as an acylphosphate, so ATP binding is part of the predicted ligase mechanism.

Reason: Both ATP-grasp folds are intact by domain signature, and the C-terminal module keeps the Ddl Mg2+/ATP-binding motif. ATP binding follows from the predicted ligase activity and is a sound domain-level inference.

Summary: Location predicted by AtSubP. Mass spectrometry places At-DDL in the chloroplast stroma, and the moss ortholog has a functional plastid transit peptide.

Reason: The chloroplast is part of the cytoplasm in GO, so the term is not false, but the protein's supported location is much more specific. Replace it with chloroplast stroma, which the proteomics evidence supports. The HDA chloroplast row is consistent with this.

Proposed replacements: chloroplast stroma

Summary: PAINT IBA from node PTN000566166, seeded by the experimentally characterized E. coli DdlB and DdlA and M. tuberculosis Ddl. The Arabidopsis protein sits inside the plant fused-DDL clade. Its moss member (Pp-DDL) is supported genetically as a D-Ala:D-Ala ligase: the knockout accumulates D-Ala, and D-Ala-D-Ala dipeptide, but not the other dipeptides tested, rescues it. Wei et al. 2020 flagged the companion peptidoglycan biosynthesis IBA on this protein, but not this MF row. Our residue check finds no Arabidopsis-specific loss of catalytic residues. The C-terminal module carries both PROSITE Ddl signatures, the active-site Glu and all three Mg2+ ligands. The N-terminal module has a diverged Mg2+ loop, but moss Pp-DDL shares the same pattern. The activity has not been assayed for the Arabidopsis protein.

Reason: To challenge an IBA, one has to show that the target lies outside the clade that inherited the function, or that the target has lost or changed it. Neither applies to the molecular function. At3g08840 groups with the plant fused DDLs, including the moss enzyme, whose ligase function is genetically supported. Its active-site and Mg2+-binding residues are conserved to the same extent as in the moss enzyme. The evidence for divergence in Arabidopsis comes from the process level: there is no PBP, the Mur pathway is incomplete, and ddl mutants have no plastid phenotype. That argues against a peptidoglycan biosynthesis role, which is why the BP IBA has rightly gone, but it does not argue against the enzyme activity. The AtMurE precedent shows that an enzymatically competent gymnosperm MurE can fulfil the repurposed Arabidopsis role, so losing the pathway does not imply losing catalysis. ACCEPT, noting that the activity has not been tested biochemically (Hirano et al. 2016).

Summary: InterPro2GO mapping from the D-Ala-D-Ala ligase C-terminal domain (IPR011095), present twice. This is consistent with the IBA row for the same term and with the conserved PROSITE Ddl signatures (PS00843, PS00844).

Reason: Same claim as the IBA row and accepted on the same grounds. The domain signature is intact, and the Mg2+-binding and active-site residues of the C-terminal module are conserved. The activity has not been confirmed biochemically for this protein.

Summary: Identified in a large-scale MS analysis of purified Arabidopsis chloroplasts, which assigned it to the stroma. Moss DDL is also plastid-targeted, and the other retained Mur-pathway homologs of Arabidopsis (e.g. AtMurE) are plastidial.

Reason: High-throughput proteomics consistent with the predicted and ortholog-supported plastid targeting. The location is correct; the more specific chloroplast stroma is proposed via the MODIFY of the cytoplasm row.

Summary: Generic metal-binding term from the ATP-grasp InterPro entry. In Ddl enzymes the metal is Mg2+, two ions per active site coordinating the ATP phosphates. The C-terminal module of At3g08840 keeps all three DdlB Mg2+-ligand positions (D881, E895, N897).

Reason: Correct but too general. The conserved ligands and the family mechanism point to Mg2+ specifically, so magnesium ion binding is the better-supported term.

Proposed replacements: magnesium ion binding