Arg (CG18104, O76895) notes

Module: dmel_arginine_metabolism (with Ass, Argl).

Literature journal

• Samson 2000 cloned the single fly arginase, nested around elav: PMID:10878001
• Insect arginase is not a urea-cycle/ammonia-detox enzyme: PMID:10878001
• Expression: PMID:10878001; mutant is viable with delay PMID:10878001
• No urea cycle in insects: PMID:40081835; arginase retained PMID:40081835
• Mitochondrial Mn enzymes listed in a fly Mn-depletion study: PMID:31799578 (general statement, not a fly localisation experiment).
• Sequence: N-terminus MWWSRKFASRSLRLHRLKST is Arg/Lys-rich with no acidic residues, resembling a mitochondrial presequence (my inspection of the UniProt sequence; no TargetP run).
• Deep research (falcon) agrees: localisation unproven; DJ-1beta PD model shows Arg/Argl transcript increase (Solana-Manrique 2022, not cached).

Decisions

• urea cycle (IEA UniPathway): REMOVE (no OTC in insects). Same call on Ass.
• ammonium excretion (ISS from Aedes): MARK_AS_OVER_ANNOTATED.
• mitochondrion ISS: UNDECIDED; cytosol IBA: UNDECIDED; cytoplasm IBA: ACCEPT (revised; see below).

Localization decision (revised)

The mitochondrion ISS row is now UNDECIDED, matching the cytosol IBA. The
presequence-like N-terminus was read from the sequence by eye, with no targeting
predictor run, and the deep research finds no fly evidence that separates a
mitochondrial from a cytosolic arginase. Cytoplasm (accepted) is used as the core
location until localization data exist.