SUCLG1 (P53597) notes

2026-11-11 first review

Written as part of the evolutionary review of the succinyl-CoA synthetase families (PTHR11117 alpha, PTHR11815 beta); see genes/yeast/LSC2/LSC2-bioinformatics/RESULTS.md.

• SUCLG1 is the single alpha subunit shared by the two human enzymes PMID:30470562.
• The alpha subunit does not set nucleotide specificity PMID:9765290. CoA binds the alpha subunit PMID:32627745.
• Loss of SUCLG1 removes both beta subunits and all activity, with rescue PMID:27484306.
• Evolutionary reading of the GDP-forming rows: accepted here. Human SUCLG1 is the case the PAINT root placement was built from; the same node over-propagates to the fungal alpha subunits in the PANTHER tree (LSC1 removes it, S. pombe marks it over-annotated). The difference is the beta repertoire of the genome, not anything in the alpha subunit. (Corrected 2026-11-12: PTN000122012 is the LUCA node of the alpha tree, not its root, and the loss of the G-type beta gene is within fungi, not kingdom-wide.)
• GO:0004774 succinate-CoA ligase activity (IMP) marked TOO_GENERAL. The paper assayed ADP-dependent and GDP-dependent activity separately, and both were lost and rescued, so the two children are supported and the parent adds nothing PMID:27484306. (Corrected 2026-11-12: this bullet used to say the row was accepted at the parent level because the assay did not separate the two enzymes. Both statements were wrong.)
• Side activities (malate-CoA ligase, acid-thiol ligase, malate metabolic process) kept as non-core, matching SUCLA2. Toxic metabolite repair was marked over-annotated in this round and changed to UNDECIDED on 2026-11-12 (see below) PMID:40108300. PMID:40108300 is abstract-only in the cache.
• RNA binding (HDA, mRNA interactome capture) marked over-annotated.
• Protein binding IPI rows with SUCLG2 removed under the protein-binding policy; the interaction is real and is the GDP-forming complex.

Affinage

Fetched with projects/AFFINAGE_EVALUATION/affinage_deep_research.py (trust gates clear). Recorded as a reference with correctness DISPUTED (changed to LOW_QUALITY on 2026-11-12, with the dispute moved to the grounding finding): the narrative claims that could be checked are right, but its GO grounding "catalytic activity, acting on a protein" is wrong for SUCLG1. Leads it supplied, none imported as annotations: restriction of POLRMT succinylation PMID:38649537, Lys90 butyrylation and SIRT4 (PMID:41991142, not cached, unverified). Both are indirect or regulatory and are recorded as suggested questions.

Not done

No falcon or perplexity deep research (keys not available in this session). Literature beyond GOA was limited to the Affinage citations and the family literature.

Retrospective audit corrections (2026-11-12)

A retrospective audit of the succinyl-CoA synthetase reviews found errors in this review. Each was checked against the source before it was changed.

• Itaconate. PMID:34492704 was cited as support for itaconate-CoA ligase activity, but it reports the opposite PMID:34492704 and that itaconate inhibits the enzyme. The quote was removed from the proposed new term. The negative result is now a finding of that reference, marked DISPUTED against PMID:40108300. The description says that itaconyl-CoA formation is reported but contested. The GO:0004775 EXP row from this paper stays ACCEPT: recombinant SCS-A did convert succinate PMID:34492704.
• PMID:40108300. Only the abstract is cached, and the published full text is not open. The bioRxiv preprint (doi 10.1101/2024.09.25.614871, version 1) was read. It assays recombinant human SUCLG1/SUCLA2 and SUCLG1/SUCLG2 with succinate, malate and itaconate; catalytic efficiencies are 30-fold (ADP-forming) and 15-fold (GDP-forming) lower with itaconate, and 340-fold and 540-fold lower with malate, than with succinate. The proposed term itaconate-CoA ligase activity is kept on that basis, with the conflict stated. The CLYBL abstract sentence was removed from the GO:0004775 and GO:0004776 EXP rows because it says nothing about succinate-CoA ligase activity. The reference stays UNVERIFIED because the published version was not read.
• GO:0110052 toxic metabolite repair (IDA, PMID:40108300): MARK_AS_OVER_ANNOTATED changed to UNDECIDED, the same as SUCLA2. The abstract and the preprint present succinyl-CoA synthetase as a source of malyl-CoA and CLYBL as the repair enzyme, which argues for over-annotation. The preprint also names succinyl-CoA synthetase with AUH and CLYBL as an enzyme of itaconate catabolism, so the curator may have annotated the activating step of a disposal route. The published paper was not read, so the experimental row is not overruled. A suggested question records what would settle it.
• GO:0004774 (IMP): action unchanged (TOO_GENERAL). The quote showing separate ADP- and GDP-dependent assays was added to this row, to the GO:0004776 EXP and IBA rows, to the GO:0004775 IBA row and to both core functions.
• GO:0016878 acid-thiol ligase (IDA, contributes_to): action unchanged (KEEP_AS_NON_CORE), reason rewritten. Malyl-CoA formation is covered by the descendant GO:0050074, which is on this gene. The row is retained only because itaconate-CoA ligase activity has no GO term. The proposed term's parent is GO:0016878 and not GO:0016405, which QuickGO defines as forming AMP and diphosphate.
• GO:0003723 RNA binding (HDA): action unchanged. The argument that the enzyme acts in the mitochondrial matrix was dropped from the reason. Table 1 of the paper (author-hosted PDF of the published article, not cached) lists SUCLG1 in the mRNA interactome class with 16 other metabolic enzymes.
• Phylogeny wording. The GO:0004776 IBA reason said fungal genomes lost the G-type beta paralog. The reworked family analysis shows the loss is within fungi [file:yeast/LSC2/LSC2-bioinformatics/RESULTS.md "The 15 fungal genomes in the PANTHER tree (14 Dikarya and the chytrid Batrachochytrium dendrobatidis) have an A-type beta gene and no G-type beta gene."], so the reason now speaks of the fungal alpha subunits in the tree. PTN000122012 is now called the LUCA node of the alpha-subunit family, not its root, on the four IBA rows that name it. Quotes of RESULTS.md sentences that no longer exist were replaced.
• Mitochondrion rows on IBA, IEA, NAS and TAS evidence: the sentence about an assay not resolving the sub-compartment was replaced by what each source rests on (the PAINT node and its seeds, the UniProt line from HAMAP rule MF_03222, an author statement). Actions unchanged.
• References. PMID:20693550: the note said the loss of SUCLA2 protein was not in the abstract. It is PMID:20693550; note corrected and finding added. Affinage record: correctness changed from DISPUTED to LOW_QUALITY, and the DISPUTED status moved to its GO grounding finding. Its one supporting quote on the GO:0004774 row was kept, beside the primary paper it was verified against. (Removed later on 2026-11-12: a record rated LOW_QUALITY should not support an annotation row; the row keeps its primary support. The validator's advisory warning that no annotation cites the deep research file is accepted.)
• Status changed from INITIALIZED to DRAFT.
• Cross-file consistency pass (2026-11-12). GO:0003824 catalytic activity (IEA) stays TOO_GENERAL; its reason is now worded the same way in all eleven subunit reviews of the family.