AASDH (ACSF4-U26) is a beta-alanine-activating enzyme. It belongs to the ATP-dependent AMP-binding (adenylating) enzyme family and has the domain architecture of a minimal non-ribosomal peptide synthetase module: an adenylation domain, a carrier domain bearing a phosphopantetheine prosthetic group, and a C-terminal PQQDH-related domain of unknown role. The enzyme adenylates beta-alanine at the expense of ATP and then transfers it onto its own phosphopantetheine arm to form a covalent acyl-enzyme thioester; a point mutant lacking the phosphopantetheine attachment site cannot form that bond. Specificity is near-absolute for beta-alanine among the standard amino acids, with a KM of about 5 micromolar. That enzymology was established on the mouse orthologue; UniProt annotates the human FUNCTION line by similarity from mouse Acsf4 (Q80WC9), and no assay of the human protein has been reported. What happens next is unknown: the physiological acceptor of the activated beta-alanine has never been identified, transfer onto small thiols can be demonstrated in vitro but was judged physiologically irrelevant by the authors who observed it, and no downstream pathway or in vivo role has been established. AASDH is ubiquitously expressed in adult tissues. It has been suggested, but not shown, that the enzyme performs an infrequent post-translational or post-transcriptional beta-alanylation.
Definition: Catalysis of the ATP-dependent activation of beta-alanine and its transfer to the phosphopantetheine prosthetic group of a carrier protein or carrier domain, forming a beta-alanyl thioester with release of AMP and diphosphate.
Justification: The most specific term currently available for AASDH is GO:0016878 acid-thiol ligase activity, which is correct but says nothing about either the substrate or the acceptor - the two facts that actually distinguish this enzyme. AASDH is the only characterised vertebrate beta-alanine-activating enzyme, its specificity is near-absolute among the standard amino acids, and the acceptor is its own phosphopantetheine arm rather than free coenzyme A, which is precisely what separates it from the acyl-CoA synthetase paralogs it is repeatedly confused with - including ACSF3, the source of the erroneous ISS removed in this review. GO already has substrate-and-carrier-specific ligase terms for other adenylating enzymes of this architecture, so the gap for beta-alanine appears historical. Note the proposed definition covers only the demonstrated half-reaction and deliberately names no downstream acceptor, since none has been identified.
Parent term: acid-thiol ligase activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0006631 fatty acid metabolic process | ISS GO_REF:0000024 | REMOVE | Summary: This is the only annotation AASDH has, and it describes a paralog's substrate rather than its own. The WITH/FROM field cites UniProtKB:Q4G176, which is ACSF3 - a human paralog, not an ortholog. ACSF3 is a genuine fatty-acid enzyme: malonyl-CoA synthetase, which catalyses the initial reaction of intramitochondrial fatty acid synthesis by activating malonate and methylmalonate. Fatty acid metabolic process is correct for ACSF3. AASDH is a different enzyme. Direct biochemistry on the recombinant protein shows it is a beta-alanine-activating enzyme, with near-absolute specificity for beta-alanine among the standard amino acids and a KM of about 5 micromolar, forming a covalent thioester through its phosphopantetheine group. Beta-alanine is a beta-amino acid, not a fatty acid, and no fatty-acid substrate has ever been reported for AASDH. What the two proteins share is the ATP-dependent AMP-binding adenylation fold and a family naming convention ("acyl-CoA synthetase family member 4" versus "member 3"), which is exactly the kind of resemblance that produces a wrong paralog transfer. Note that the same assumption has propagated into UniProt's keyword set, which lists "Fatty acid metabolism" and "Lipid metabolism" for AASDH even though UniProt's own FUNCTION line describes beta-alanine activation. That is worth correcting upstream too. Reason: A sequence-similarity inference from a paralog, contradicted by direct enzymology on the gene itself. This is a demonstrably wrong ISS argued on biological grounds, not the second-guessing of an experimental annotation. Propagation Review Root cause: PROPAGATION BAD Failure modes: WRONG ORTHOLOG OR PARALOG FUNCTIONAL DIVERGENCE Sources checked: UniProtKB:Q4G176 ACSF3, a human paralog rather than an ortholog. It is a genuine malonyl-CoA synthetase acting in intramitochondrial fatty acid synthesis, so the term is right for the source and wrong for the target; the two share only the adenylating-enzyme fold and an ACSF family label. Supporting Evidence: PMID:24467666 We conclude that ACSF4-U26 is a Ξ²-alanine-activating enzyme PMID:24467666 Competition experiments with various amino acids indicated that the reaction was almost specific for Ξ²-alanine, and a KM of ~ 5 ΞΌm was calculated for this reaction. file:human/AASDH/AASDH-uniprot.txt CC -!- FUNCTION: Covalently binds beta-alanine in an ATP-dependent manner to |
| GO:0016878 acid-thiol ligase activity | ISS PMID:24467666 Vertebrate Acyl CoA synthetase family member 4 (ACSF4-U26) i... | NEW | Summary: Proposed new annotation, and the core molecular function. Removing the incorrect fatty-acid ISS leaves AASDH with no annotations at all, yet its chemistry is directly demonstrated: in the presence of ATP, purified recombinant enzyme progressively formed a covalent bond with radiolabelled beta-alanine, and the bond was abolished in a point mutant lacking the phosphopantetheine attachment site. That is ATP-dependent formation of a thioester bond between a carboxylic acid and a thiol, which is what acid-thiol ligase activity denotes. The affinage grounding proposed GO:0016874 ligase activity and GO:0140657 ATP-dependent activity, both correct but two levels too general; GO:0016878 is the specific descendant the evidence supports. The assay is on the MOUSE protein ("purified mouse recombinant ACSF4-U26"), and UniProt annotates the human FUNCTION line as ECO:0000250|UniProtKB:Q80WC9, so the human annotation is an ortholog transfer. Reason: Supplies the molecular function the gene entirely lacks, at the level of specificity the enzymology supports. Evidence code is ISS with UniProtKB:Q80WC9 (mouse Acsf4) as the supporting entity, not IDA: the experiments are on the mouse protein. This is deliberate given what this review argues - it REMOVES an ISS for naming the wrong source (the paralog ACSF3/Q4G176), so asserting a direct human assay for evidence that is itself an ortholog transfer would apply the standard in only one direction. A genuine ortholog transfer with the correct source is exactly the contrast case. Further support: UniProt assigns EC=6.2.1.-, and 6.2.1 is the acid-thiol ligase subclass this term denotes. Supporting Evidence: PMID:24467666 In the presence of ATP, purified mouse recombinant ACSF4-U26 progressively formed a covalent bond with radiolabelled Ξ²-alanine. PMID:24467666 The bond was not formed in a point mutant lacking the phosphopantetheine attachment site. file:human/AASDH/AASDH-uniprot.txt DE EC=6.2.1.-; |
| GO:0031177 phosphopantetheine binding | ISS PMID:24467666 Vertebrate Acyl CoA synthetase family member 4 (ACSF4-U26) i... | NEW | Summary: Proposed new annotation. AASDH carries a carrier domain (UniProt residues 553-630) whose phosphopantetheine prosthetic group is the acceptor for the activated beta-alanine, and the requirement is demonstrated rather than inferred: a point mutant lacking the phosphopantetheine attachment site fails to form the thioester bond. This is a distinct molecular function from the ligase activity and is what makes AASDH a carrier-protein enzyme of the non-ribosomal peptide synthetase type rather than a simple acyl-CoA synthetase. Reason: Captures the prosthetic-group requirement that defines this enzyme's mechanism and distinguishes it from the acyl-CoA synthetase paralogs it is confused with. ISS with UniProtKB:Q80WC9 for the same reason as GO:0016878 - the point-mutant experiment is on the mouse protein. Supporting Evidence: PMID:24467666 The bond was not formed in a point mutant lacking the phosphopantetheine attachment site. file:human/AASDH/AASDH-uniprot.txt FT DOMAIN 553..630 file:human/AASDH/AASDH-deep-research-affinage.md Catalysis depends on a phosphopantetheine cofactor: a point mutant lacking the phosphopantetheine attachment site fails to form the thioester bond |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: What is the physiological acceptor of the activated beta-alanine? This is the central gap. The enzyme demonstrably loads beta-alanine onto its phosphopantetheine arm, but the only unloading observed in vitro was onto cysteine and other small thiols, which the authors judged physiologically irrelevant. Until the true acceptor is identified, no pathway or downstream process annotation can be justified, and the hypothesis of a rare post-translational or post-transcriptional beta-alanylation remains untested.
Q: What does the C-terminal PQQDH-related domain do? Its deletion does not abolish beta-alanine transfer onto thiols, so the adenylation domain carries that activity, but the domain is conserved and its function is uncharacterised. It is a plausible candidate for acceptor recognition, which would make it the route to answering the question above.
Q: Should UniProt's "Fatty acid metabolism" and "Lipid metabolism" keywords for AASDH be withdrawn? They appear to derive from the same paralog-based assumption as the GO ISS removed in this review, and they conflict with UniProt's own FUNCTION line describing beta-alanine activation.
Q: Does AASDH have a role in carnosine or pantothenate biology? Beta-alanine in vertebrates is principally consumed by carnosine synthesis and is a precursor of coenzyme A via pantothenate. Neither connection has been tested for AASDH, but they are the obvious places a beta-alanine-activating enzyme might act.
Experiment: Incubate loaded recombinant AASDH with cell lysate fractions and identify beta-alanylated products by mass spectrometry using isotopically labelled beta-alanine, separating protein, RNA and small-molecule acceptor fractions to test the post-translational versus post-transcriptional hypotheses directly.
Hypothesis: The activated beta-alanine is transferred to a specific macromolecular acceptor rather than to free thiols.
Type: biochemistry and mass spectrometry
Experiment: Compare full-length AASDH with a PQQDH-domain deletion for acceptor usage in the assay above, and use the isolated domain as bait in pull-downs to identify what it binds. A domain that is dispensable for thiol transfer but required for a specific acceptor would identify both the domain's role and the acceptor.
Hypothesis: The C-terminal PQQDH-related domain selects the physiological acceptor.
Type: structure-function
Experiment: Knock out AASDH in a human cell line and profile beta-alanine, carnosine, pantothenate and coenzyme A pools by targeted metabolomics, with isotope tracing from labelled beta-alanine to establish whether AASDH flux is quantitatively significant in any of these routes.
Hypothesis: Loss of AASDH perturbs beta-alanine-dependent metabolism.
Type: metabolomics
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)