YciO is an uncharacterized member of the PF01300 (TsaC/Sua5/YciO/YrdC) family in E. coli. It is a PARALOG of TsaC (L-threonylcarbamoyladenylate synthase, EC 2.7.7.87), but extensive evidence demonstrates that YciO does NOT perform the same biological function as TsaC. While YciO shows very weak L-threonylcarbamoyladenylate synthase activity in vitro (0.14 nM/min TC-AMP production, PMID:37963869), this is more than 4 orders of magnitude weaker than TsaC (2.8 uM/min) under similar conditions (PMID:40703034). The canonical KRSN tetrad (KxR...SxN) required for TC-AMP synthesis is replaced by KxL...SxM in YciO, consistent with altered or lost catalytic activity. In vivo experiments have shown that TsaC is essential even when yciO is present or overexpressed, indicating YciO cannot substitute for TsaC function. The weak in vitro activity is likely residual ancestral/promiscuous activity, not the biological function. YciO possesses a large conserved positively charged surface (absent in TsaC) predicted to interact with RNA (PMID:40703034). Genomic context analysis shows yciO genes frequently colocalize with rnm genes encoding RNase AM, a 5-to-3 exonuclease that matures the 5-prime ends of rRNAs. The true biological function of YciO is unknown but likely relates to rRNA metabolism rather than tRNA modification. This gene is a textbook case of how paralog confusion and uncritical acceptance of in vitro promiscuous activity can lead to incorrect functional annotations.
Summary: IBA annotation for cytosol localization based on phylogenetic inference (PANTHER). Consistent with experimental IDA evidence from two independent proteomics studies (PMID:15911532, PMID:18304323) that identified YciO in the cytosolic fraction of E. coli K-12.
Reason: Cytosolic localization is well-supported by both phylogenetic inference and direct experimental evidence from mass spectrometry-based proteomics. Two independent studies identified YciO as a cytosolic protein.
Here we describe a proteomic analysis of Escherichia coli in which 3,199 protein forms were detected, and of those 2,160 were annotated and assigned to the cytosol, periplasm, inner membrane, and outer membrane by biochemical fractionation followed by two-dimensional gel electrophoresis and tandem mass spectrometry.
Summary: IEA annotation based on InterPro domain IPR006070 (Sua5-like domain) mapping to dsRNA binding. This is a computational prediction from InterPro. While YciO does possess a large conserved positively charged surface predicted to interact with RNA (PMID:40703034), the specific annotation as dsRNA binding is not experimentally validated and the InterPro mapping may be overly specific. The positively charged surface is a feature distinguishing YciO from TsaC, suggesting RNA interaction is plausible, but the specific type of RNA (dsRNA vs rRNA) is uncertain.
Reason: The InterPro-based prediction of dsRNA binding is plausible given the large positively charged surface unique to YciO proteins (PMID:40703034), but there is no direct experimental evidence for dsRNA binding specifically. The genomic context (colocalization with rnm genes encoding RNase AM, a rRNA maturation enzyme) suggests rRNA interaction may be more likely. Without experimental evidence, it is difficult to determine whether dsRNA binding or a more specific RNA binding term is appropriate. The annotation should be revisited when experimental data on YciO RNA binding become available.
genes are colocalized with rnm genes (Fig. 7), which encode the recently characterized RNase AM, a 5β² to 3β² exonuclease that matures the 5β² end of all 3 ribosomal RNAs in E. coli
Summary: IEA annotation based on EC number mapping (EC:2.7.7.87). This EC number was assigned to YciO in UniProt based on the Kim et al. 2023 deep learning prediction and in vitro assay (PMID:37963869). However, de Crecy-Lagard et al. 2025 (PMID:40703034) demonstrated that this is a classic case of paralog incorrect (PLI) annotation. YciO is a paralog of TsaC but does not perform the same biological function. The in vitro activity is 4 orders of magnitude weaker than TsaC and represents residual ancestral/promiscuous activity. In vivo experiments show TsaC is essential even with yciO present. The EC2GO mapping propagates this incorrect EC assignment.
Reason: This annotation propagates through EC2GO from the incorrect EC:2.7.7.87 assignment to YciO. De Crecy-Lagard et al. 2025 (PMID:40703034) provide compelling evidence that YciO does NOT function as an L-threonylcarbamoyladenylate synthase in vivo. The weak in vitro activity (0.14 nM/min vs 2.8 uM/min for TsaC) is residual ancestral activity, not the biological function. In vivo, TsaC is essential even when yciO is overexpressed. The EC number 2.7.7.87 prediction was given a confidence score of 0 (refuted) by expert curation in the de Crecy-Lagard study.
the activity reported (0.14 nM/min TC-AMP production rate) for E. coli YciO is more than 4 orders of magnitude weaker than that of E. coli TsaC (2.8 ΞΌM/min) at the same enzyme concentration and similar reaction conditions
the functional puzzle is far from being solved for proteins of the YciO subgroup, and even if the existing data suggest a role in RNA metabolism, it cannot be the same as TsaC, and the EC number 2.7.7.87 prediction was given a CS of 0
IDA PMID:37963869 Functional annotation of enzyme-encoding genes using deep le...
REMOVE
Summary: IDA annotation based on Kim et al. 2023 (PMID:37963869), which used DeepECTransformer to predict EC 2.7.7.87 for YciO and validated it with an in vitro enzyme assay showing a specific activity of 0.0705 U/mg. However, de Crecy-Lagard et al. 2025 (PMID:40703034) demonstrated this is a paralog incorrect (PLI) prediction. The measured activity (0.14 nM/min TC-AMP) is more than 4 orders of magnitude weaker than genuine TsaC activity (2.8 uM/min). This weak activity represents residual ancestral/promiscuous catalytic activity from shared evolutionary origin, not the biological function. In vivo experiments show YciO cannot substitute for TsaC. Models of enzyme evolution predict that paralogs retain promiscuous ancestral activities detectable in vitro. The IDA evidence code is technically correct for in vitro detection, but the annotation misrepresents the biological function of YciO.
Reason: While Kim et al. 2023 (PMID:37963869) did detect L-threonylcarbamoyladenylate synthase activity in vitro, de Crecy-Lagard et al. 2025 (PMID:40703034) conclusively demonstrate this is promiscuous/residual ancestral activity, not the biological function. The activity is 4 orders of magnitude weaker than TsaC. In vivo, TsaC is essential even with YciO present or overexpressed. The Kim et al. study used the in vitro assay to validate a deep learning prediction without considering the in vivo context, genomic evidence, or the distinction between promiscuous and biological activity. As stated in PMID:40703034: in vitro activity alone is not sufficient to validate protein function in vivo. This annotation should be removed as it represents an over-annotation based on promiscuous activity of a paralog.
In the case of YciO, which was previously annotated to belong to the SUA5 family, DeepECtransformer predicted its EC number to be EC:2.7.7.87 (L-threonylcarbamoyladenylate synthase) with the prediction score of 0.9108. The specific activity of YciO was measured to be 0.0705 U mg-1
the activity reported (0.14 nM/min TC-AMP production rate) for E. coli YciO is more than 4 orders of magnitude weaker than that of E. coli TsaC (2.8 ΞΌM/min) at the same enzyme concentration and similar reaction conditions ... consistent with the possibility of a missing partner or a different biological substrate for YciO
TsaC is an enzyme predicted to have been present in the Last Universal Common Ancestor ... YciO is a likely paralog of TsaC (Fig. 6), and as such, it is likely to have residual ancestral catalytic activity that can be detected in vitro
in vitro activity alone is not sufficient to validate the function of a protein in vivo
file:ECOLI/yciO/yciO-deep-research-falcon.md
Falcon deep research independently confirms YciO is not part of the canonical t6A pathway. Multiple sources (Harris 2015, Thiaville 2014, Carvalho 2017, Pichard-Kostuch 2023) establish YciO as a divergent paralog lacking the conserved KRSN tetrad required for TC-AMP synthesis.
IDA PMID:15911532 Localization, annotation, and comparison of the Escherichia ...
ACCEPT
Summary: IDA annotation for cytosol localization from Lopez-Campistrous et al. 2005 (PMID:15911532), a large-scale proteomics study of E. coli K-12 that assigned proteins to subcellular compartments by biochemical fractionation followed by 2D gel electrophoresis and tandem mass spectrometry.
Reason: Direct experimental evidence from proteomics-based subcellular fractionation. The study identified 2,160 protein forms and assigned them to cytosol, periplasm, inner membrane, and outer membrane. YciO is a soluble cytoplasmic protein consistent with this localization.
Here we describe a proteomic analysis of Escherichia coli in which 3,199 protein forms were detected, and of those 2,160 were annotated and assigned to the cytosol, periplasm, inner membrane, and outer membrane by biochemical fractionation followed by two-dimensional gel electrophoresis and tandem mass spectrometry.
IDA PMID:18304323 Protein abundance profiling of the Escherichia coli cytosol.
ACCEPT
Summary: IDA annotation for cytosol localization from Ishihama et al. 2008 (PMID:18304323), a comprehensive protein abundance profiling study of the E. coli cytosol using mass spectrometry.
Reason: Independent experimental confirmation of cytosolic localization from a second large-scale proteomics study. Consistent with the IDA annotation from PMID:15911532 and the IBA annotation.
ISS PMID:40703034 Limitations of current machine learning models in predicting...
NEW
Summary: NEW annotation. YciO possesses a large conserved positively charged surface (comprising ~6% of total molecular surface area) that is present and conserved in all YciO family members but absent in TsaC proteins (PMID:40703034). This surface is predicted to interact with RNA. Genomic context shows yciO genes colocalize with rnm genes (encoding RNase AM, a rRNA maturation enzyme), suggesting involvement in rRNA metabolism.
Reason: De Crecy-Lagard et al. 2025 (PMID:40703034) provide structural and genomic context evidence that YciO has a conserved positively charged surface predicted to bind RNA. While the specific RNA target has not been experimentally determined, general RNA binding is well supported by structural analysis. This is proposed as a more appropriate molecular function annotation than the incorrect L-threonylcarbamoyladenylate synthase activity.
This large positively charged surface is conserved in YciO proteins and is absent in TsaC proteins
Core Functions
YciO is a protein of unknown biological function that likely participates in rRNA metabolism based on genomic context (colocalization with rnm/RNase AM genes) and structural features (conserved positively charged surface predicted to bind RNA). It is NOT an L-threonylcarbamoyladenylate synthase despite weak in vitro activity, which represents residual ancestral activity from its evolutionary relationship to TsaC.
the structure of YciO exhibits a large positively charged surface predicted to interact with RNA ... This large positively charged surface is conserved in YciO proteins and is absent in TsaC proteins
Q: Does YciO physically interact with RNase AM (Rnm) or other rRNA processing factors?
Q: Should the UniProt record for YciO (P0AFR4) be corrected to remove the EC 2.7.7.87 assignment?
Suggested Experiments
Experiment: RNA co-immunoprecipitation or CLIP-seq to identify RNA binding partners of YciO. YciO has a conserved positively charged surface predicted to bind RNA. Identifying the actual RNA targets (rRNA, tRNA, mRNA, or other) would resolve the functional puzzle.
Hypothesis: YciO binds rRNA based on its conserved positively charged surface and genomic colocalization with rnm (RNase AM) genes.
Experiment: In vivo complementation tests with yciO deletion in combination with rnm deletion. Genomic context suggests functional coupling between YciO and RNase AM. A synthetic interaction would support the hypothesis that YciO participates in rRNA maturation.
Hypothesis: YciO and RNase AM (Rnm) function in the same rRNA maturation pathway, and double deletion will show a synthetic phenotype.
Experiment: Electrophoretic mobility shift assays (EMSAs) with various RNA substrates including rRNA fragments, tRNA, and dsRNA to narrow down the RNA binding specificity suggested by the positively charged surface.
Hypothesis: YciO preferentially binds rRNA over other RNA species.
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.
YciO DeepECTF prediction review. The DeepECTF prediction of L-threonylcarbamoyladenylate synthase (EC 2.7.7.87) is incorrect - a classic paralog overannotation error. YciO is a paralog of TsaC but shows 10,000x weaker activity and does not complement tsaC mutants in vivo. Structural and genomic context evidence suggests a distinct role in rRNA metabolism.
Review score: 2 = concordant with evidence; 1 = uncertain; 0 = discordant with evidence. This is an assessment score, not a model probability.
Review rationale: Paralog incorrect. YciO is a member of the same Pfam family (PF01300) as TsaC/Sua5, and DeepECTF assigns the same EC number. However: (1) In vitro activity is 10,000x weaker than TsaC (0.14 nM/min vs 2.8 uM/min). (2) YciO does not complement tsaC deletion in vivo. (3) YciO has a conserved positively charged surface patch absent in TsaC, suggesting RNA-binding function. (4) yciO genes co-localize with rnm (RNase AM) genes, suggesting a role in rRNA metabolism rather than tRNA modification. The weak in vitro activity likely represents residual ancestral catalytic activity from TsaC, not the evolved biological function.
Supporting Evidence:
PMID:40703034: "the activity reported (0.14 nM/min TC-AMP production rate) for E. coli YciO is more than 4 orders of magnitude weaker than that of E. coli TsaC (2.8 ΞΌM/min)"
PMID:40703034: "YciO does not perform the same function as TsaC/Susa5 in vivo experiments"
PMID:40703034: "the structure of YciO exhibits a large positively charged surface predicted to interact with RNA...This large positively charged surface is conserved in YciO proteins and is absent in TsaC proteins"