YjhQ is an uncharacterized protein in E. coli K12 that contains a GNAT (GCN5-related N-acetyltransferase) domain (Pfam PF00583, PROSITE GNAT domain PS51186). UniProt designates it as "Uncharacterized N-acetyltransferase YjhQ" with an incomplete EC number (EC 2.3.1.-), reflecting that while the structural fold strongly suggests N-acetyltransferase activity, the specific substrate has never been experimentally determined. Yamaguchi and Inouye (PMID:26553797) identified yjhQ as the antitoxin component of a novel toxin-antitoxin (TA) system yjhX-yjhQ (renamed topAI-yjhQ), where the toxin TopAI (YjhX) inhibits topoisomerase I. YjhQ co-expression neutralizes TopAI toxicity, and the two proteins physically interact in pull-down assays. However, whether the antitoxin function of YjhQ is related to its acetyltransferase fold or represents a separate binding/sequestration activity is unknown. De Crecy-Lagard et al. 2025 (PMID:40703034) highlighted yjhQ as a case where DeepECTF incorrectly predicted EC 2.3.1.189 (mycothiol synthase), a biologically impossible assignment because mycothiol biosynthesis is entirely absent from E. coli (an Actinobacteria-specific pathway, BioCyc PWY1G-0). This represents a Non-Paralog Incorrect (NPI) prediction error arising from ML models lacking organism/pathway context.
Existing Annotations Review
GO Term
Evidence
Action
Reason
GO:0016747 acyltransferase activity, transferring groups other than amino-acyl groups
IBA GO_REF:0000033
ACCEPT
Summary: IBA annotation for acyltransferase activity based on phylogenetic inference (PANTHER family including P0A944, P46854, P9WJM7, Q9I1K2, I6YG32). YjhQ contains a well-defined GNAT domain (Pfam PF00583, InterPro IPR000182, PROSITE PS51186) and belongs to the acetyltransferase family per UniProt. The structural evidence strongly supports general acyltransferase activity at the fold level. However, the specific substrate is completely unknown. Note that the IBA inference includes P9WJM7 from Mycobacterium tuberculosis; de Crecy-Lagard et al. 2025 (PMID:40703034) showed that the specific prediction of EC 2.3.1.189 (mycothiol synthase) by DeepECTF for yjhQ is biologically impossible since mycothiol biosynthesis (BioCyc PWY1G-0) is absent from E. coli. The IBA term GO:0016747 is appropriately general and does NOT make a substrate-specific claim, so it remains valid as a domain-level prediction.
Reason: The GNAT domain fold is unambiguous from sequence and structural analysis (InterPro IPR000182, Pfam PF00583, CDD cd04301 NAT_SF, Gene3D 3.40.630.30, SUPFAM SSF55729). The IBA annotation at GO:0016747 is appropriately broad and does not commit to a specific substrate. This is a reasonable annotation for an uncharacterized acetyltransferase. The phylogenetic inference is sound at this level of generality even though the family includes mycobacterial members with substrate specificities that are impossible in E. coli.
YjhQ/b4307 is predicted to be a mycothiol synthase (EC 2.3.1.189), but mycothiol is not a molecule synthesized by E. coli, and the remaining pathway genes are absent from the genome (BioCyc ID: PWY1G-0)
file:ECOLI/yjhQ/yjhQ-deep-research-falcon.md
Falcon deep research confirms no substrate-level enzymatic reaction has been validated for YjhQ. Additional context includes biofilm-associated upregulation (expression ratio 11.3), Thr11 phosphorylation linked to YeaG kinase, and inclusion in ASKA acetyltransferase screen without positive result.
GO:0016747 acyltransferase activity, transferring groups other than amino-acyl groups
IEA GO_REF:0000002
ACCEPT
Summary: IEA annotation based on InterPro2GO mapping from InterPro domain IPR000182 (GNAT domain). This is consistent with the IBA annotation and the UniProt classification of YjhQ as an uncharacterized N-acetyltransferase (EC 2.3.1.-). The InterPro domain match is clear and unambiguous.
Reason: Straightforward domain-based annotation. The GNAT domain (IPR000182) maps to acyltransferase activity, which is appropriate for the structural fold. Redundant with the IBA annotation but independently derived and valid.
IPI PMID:26553797 An endogenous protein inhibitor, YjhX (TopAI), for topoisome...
MODIFY
Summary: IPI annotation for protein binding based on the interaction between YjhQ and TopAI (YjhX, UniProtKB:Q2EEU2) demonstrated by pull-down assay in PMID:26553797. Yamaguchi and Inouye showed that YjhQ is the antitoxin component of the topAI-yjhQ TA system, and that YjhQ physically interacts with TopAI to neutralize its toxicity. However, GO:0005515 (protein binding) is an uninformative term per GO curation guidelines. The actual function demonstrated is toxin sequestration: YjhQ binds TopAI and neutralizes its inhibition of topoisomerase I. GO:0097351 (toxin sequestering activity) would be a much more informative and specific annotation for this interaction.
Reason: Per GO curation best practices, GO:0005515 (protein binding) is discouraged as it provides no information about the nature or consequence of the binding. The experimental evidence in PMID:26553797 clearly demonstrates that YjhQ functions as an antitoxin that sequesters TopAI. GO:0097351 (toxin sequestering activity) precisely captures this molecular function. The pull-down assay and co-expression rescue experiments establish that YjhQ binding to TopAI neutralizes TopAI toxicity.
we found a novel, previously unidentified TA system in Escherichia coli named yjhX-yjhQ. Induction of YjhX (85 amino acid residues) causes cell-growth arrest resulting in cell death, while YjhQ (181 residues) co-induction resumes cell growth
co-induction of YjhQ in the presence of 0.2% arabinose with TopAI in the presence of 0.1 mM IPTG neutralized the toxicity of TopAI indicating that the overproduction of TopAI is toxic to the cells, while YjhQ is the antitoxin for TopAI
IPI PMID:26553797 An endogenous protein inhibitor, YjhX (TopAI), for topoisome...
NEW
Summary: NEW annotation proposed to replace GO:0005515. YjhQ functions as the antitoxin in the topAI-yjhQ toxin-antitoxin system, binding and sequestering TopAI to neutralize its inhibition of topoisomerase I (PMID:26553797). This is demonstrated by pull-down assays showing physical interaction and co-expression experiments showing rescue from TopAI toxicity.
Reason: GO:0097351 (toxin sequestering activity) is the appropriate molecular function term for the antitoxin role of YjhQ. The evidence from PMID:26553797 shows that YjhQ physically binds TopAI in pull-down assays, co-expression of YjhQ neutralizes TopAI-mediated cell death, and this represents a bona fide TA system. This term accurately captures the molecular mechanism by which YjhQ functions as an antitoxin.
we found a novel, previously unidentified TA system in Escherichia coli named yjhX-yjhQ. Induction of YjhX (85 amino acid residues) causes cell-growth arrest resulting in cell death, while YjhQ (181 residues) co-induction resumes cell growth
Core Functions
Predicted N-acetyltransferase activity (EC 2.3.1.-) based on GNAT domain (Pfam PF00583). Specific substrate unknown.
Antitoxin function: sequesters TopAI (YjhX) toxin to neutralize its inhibition of topoisomerase I in the topAI-yjhQ TA system (PMID:26553797). Whether this antitoxin function is mechanistically related to the acetyltransferase fold is unknown.
we found a novel, previously unidentified TA system in Escherichia coli named yjhX-yjhQ. Induction of YjhX (85 amino acid residues) causes cell-growth arrest resulting in cell death, while YjhQ (181 residues) co-induction resumes cell growth
Q: What is the specific enzymatic substrate of YjhQ N-acetyltransferase activity? Is the acetyltransferase activity required for antitoxin function, or does YjhQ neutralize TopAI purely through protein-protein binding/sequestration?
Q: Is the topAI-yjhQ TA system active under physiological conditions, or only when artificially overexpressed? What are the native expression conditions?
Suggested Experiments
Experiment: Enzymatic assay with purified YjhQ against a panel of candidate acetyl-CoA acceptor substrates (amino sugars, amino acids, antibiotics, etc.) to identify the native substrate.
Type: biochemical assay
Experiment: Structure determination of YjhQ alone and in complex with TopAI to understand whether the GNAT active site is involved in the antitoxin mechanism.
Type: structural biology
Experiment: Mutagenesis of GNAT catalytic residues to determine whether acetyltransferase activity is required for antitoxin function or whether binding alone suffices.
Type: site-directed mutagenesis
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
YjhQ DeepECTF prediction review. The DeepECTF prediction of mycothiol synthase (EC 2.3.1.189) is incorrect. Mycothiol is not synthesized by E. coli, and the remaining pathway enzymes are absent from the genome. YjhQ is actually an antitoxin component of the TopAI-YjhQ toxin-antitoxin system.
Review score: 2 = concordant with evidence; 1 = uncertain; 0 = discordant with evidence. This is an assessment score, not a model probability.
Review rationale: Nonparalog incorrect. Mycothiol is not a molecule synthesized by E. coli. The mycothiol biosynthesis pathway (BioCyc ID: PWY1G-0) is absent from E. coli - none of the other pathway enzymes are encoded in the genome. Mycothiol is produced by Actinobacteria (e.g., Mycobacterium). YjhQ is actually an N-acetyltransferase that functions as an antitoxin in the TopAI-YjhQ type IV toxin-antitoxin system, sequestering the TopAI toxin to prevent translation inhibition.
Supporting Evidence:
PMID:40703034: "YjhQ/b4307 is predicted to be a mycothiol synthase (EC 2.3.1.189), but mycothiol is not a molecule synthesized by E. coli, and the remaining pathway genes are absent from the genome (BioCyc ID: PWY1G-0)"