YjdM is a small (111 aa) uncharacterized protein in E. coli K-12, located upstream of the phnCDEFGHIJKLMNOP methylphosphonate catabolism operon. It was originally designated phnA based on its genomic position (PMID:2155230), but genetic analyses showed that disruption of yjdM has no effect on phosphonate metabolism (PMID:8335257). Kim et al. 2023 (PMID:37963869) used a deep learning tool (DeepECtransformer) to predict EC 3.11.1.2 (phosphonoacetate hydrolase) and validated this in vitro with purified recombinant protein. However, de Crecy-Lagard et al. 2025 (PMID:40703034) critically evaluated this annotation, noting that (1) the experimentally validated PhnA family is nonhomologous to YjdM, (2) expression of true PhnA family members in E. coli suggested PhnA activity was NOT endogenously present, (3) the initial report linking yjdM to phosphonate catabolism was refuted by Metcalf and Wanner 1993, and (4) genome neighborhood analysis shows yjdM homologs are NOT near phosphonate genes outside E. coli. The in vitro activity likely does not reflect the biological function. The true function of YjdM remains unknown. The protein belongs to the YjdM family (Pfam PF03831) and contains a zinc ribbon domain (Pfam PF08274).
Summary: IBA annotation for cytosol localization based on phylogenetic inference (PANTHER). This is consistent with the two IDA proteomics-based annotations (PMID:15911532, PMID:18304323) that detected YjdM in the cytosolic fraction of E. coli.
Reason: Cytosol localization is well supported by independent proteomics experiments and phylogenetic inference. Two large-scale proteomics studies detected YjdM in the cytosolic fraction (PMID:15911532, PMID:18304323), and the IBA annotation is concordant.
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 EC number mapping (EC:3.11.1.2). The EC number assignment in UniProt derives from the Kim et al. 2023 in vitro assay (PMID:37963869). However, de Crecy-Lagard et al. 2025 (PMID:40703034) critically re-evaluated this and concluded that the in vitro activity does not reflect the biological function. The prediction confidence score was low (CS=1), YjdM belongs to a nonhomologous family from the known PhnA enzymes, disruption of yjdM has no effect on phosphonate metabolism (PMID:8335257), and genome neighborhood analysis does not support a role in phosphonate catabolism outside E. coli.
Reason: This IEA annotation transitively inherits the problematic EC number assignment. The EC:3.11.1.2 assignment to YjdM is based solely on in vitro activity from a deep learning prediction (PMID:37963869) that has been critically challenged by de Crecy-Lagard et al. 2025 (PMID:40703034). Multiple lines of evidence argue against this being the biological function: (1) disruption of yjdM has no effect on phosphonate metabolism in E. coli B strains (PMID:8335257), (2) the true PhnA family is nonhomologous, (3) genome neighborhoods show yjdM is not near phosphonate genes outside E. coli, and (4) expression of true PhnA family members in E. coli suggested this activity was not endogenously present. This is a clear case where in vitro activity does not equal in vivo function.
YjdM was predicted by DeepECtransformer to have the EC number EC:3.11.1.2 (phosphonoacetate hydrolase). The specific phosphonoacetate hydrolase activity of YjdM obtained by enzyme assay was 139.85 U mg-1
the initial report that the protein was involved in phosphonate catabolism was later refuted with additional genetic analyses (Metcalf and Wanner 1993)
IDA PMID:37963869 Functional annotation of enzyme-encoding genes using deep le...
MARK AS OVER ANNOTATED
Summary: IDA annotation based on Kim et al. 2023 who showed in vitro phosphonoacetate hydrolase activity (139.85 U/mg) for purified recombinant YjdM. The protein was overexpressed in E. coli BL21(DE3), purified via His-tag, and assayed with phosphonoacetic acid as substrate measuring phosphate release. However, this was guided by a deep learning prediction (DeepECtransformer, prediction score 0.6103) and the training sequence with highest similarity to YjdM had a different EC number (EC:2.7.7.6). De Crecy-Lagard et al. 2025 (PMID:40703034) challenged this annotation, noting low confidence, nonhomology with true PhnA family, genetic refutation of phosphonate catabolism role (PMID:8335257), and lack of genome neighborhood support.
Reason: While the in vitro enzymatic activity was demonstrated experimentally (PMID:37963869), multiple lines of evidence strongly suggest this does not represent the biological function of YjdM. (1) Metcalf and Wanner 1993 showed disruption of yjdM has no effect on phosphonate metabolism (PMID:8335257). (2) YjdM is nonhomologous to the true PhnA phosphonoacetate hydrolase family. (3) Genome neighborhood analysis shows yjdM homologs are not near phosphonate genes outside E. coli (PMID:40703034). (4) The DeepECtransformer prediction had a low score (0.6103) and the most similar training sequence had a different EC number. (5) Expression of true PhnA family members in E. coli suggested phosphonoacetate hydrolase activity was not endogenously present. This is a paradigmatic case of in vitro activity not equaling in vivo function. Promiscuous in vitro hydrolysis of a C-P bond does not constitute evidence for biological function. The IDA evidence code is technically correct for the in vitro observation but misleading as a functional annotation.
YjdM was predicted by DeepECtransformer to have the EC number EC:3.11.1.2 (phosphonoacetate hydrolase). The specific phosphonoacetate hydrolase activity of YjdM obtained by enzyme assay was 139.85 U mg-1
for YjdM, predicted by the neural network as EC:3.11.1.2 with a prediction score of 0.6103, the training sequence with the highest sequence similarity had a different EC number (C9Y1B8_CROTZ; EC:2.7.7.6)
Underscoring the challenges in annotations, the initial report that the protein was involved in phosphonate catabolism was later refuted with additional genetic analyses (Metcalf and Wanner 1993)
expression of members of this family in E. coli suggested that PhnA activity was not present in this organism (Kulakova et al. 1997)
file:ECOLI/yjdM/yjdM-deep-research-falcon.md
Falcon deep research confirms in vitro activity (139.85 U/mg) but highlights that E. coli yjdM is not embedded in a typical phosphonoacetate utilization module. E. coli primarily uses C-P lyase for phosphonate catabolism, and phnA (yjdM) was shown to be non-essential for phosphonate metabolism.
IDA PMID:15911532 Localization, annotation, and comparison of the Escherichia ...
ACCEPT
Summary: IDA annotation based on Lopez-Campistrous et al. 2005, a large-scale proteomics study that fractionated E. coli K-12 into subcellular compartments and detected proteins by 2D-GE and tandem mass spectrometry. YjdM was identified in the cytosolic fraction.
Reason: Cytosol localization determined by biochemical fractionation and mass spectrometry in a large-scale proteomics study. This is a standard high-quality method for determining subcellular localization in bacteria. Consistent with the IBA annotation and the independent proteomics study in PMID:18304323.
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 based on Ishihama et al. 2008, a comprehensive proteomics study of the E. coli cytosol using LC-MS/MS with multiple fractionation approaches. YjdM was among the 1103 proteins identified from the cytosolic fraction.
Reason: Cytosol localization determined by mass spectrometry-based proteomics with extensive fractionation. Independent confirmation of the PMID:15911532 result. Consistent with the IBA annotation.
we identified 1103 proteins from the cytosolic fraction of the Escherichia coli strain MC4100
Core Functions
The biological function of YjdM remains unknown. It is a cytosolic protein of the YjdM family (IPR004624) with a zinc ribbon domain, but its molecular activity and biological role have not been established. The phosphonoacetate hydrolase activity observed in vitro (PMID:37963869) is not supported by in vivo evidence and has been critically challenged (PMID:40703034).
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.
YjdM DeepECTF prediction review. The DeepECTF prediction of phosphonoacetate hydrolase (EC 3.11.1.2) was validated in vitro but is uncertain for in vivo function. The in vitro PhnA activity may not represent the biological function - genome neighborhood analysis shows yjdM genes are generally not near phosphonate catabolism genes except in E. coli, and the experimentally validated PhnA belongs to a nonhomologous family.
Review score: 2 = concordant with evidence; 1 = uncertain; 0 = discordant with evidence. This is an assessment score, not a model probability.
Review rationale: Uncertain. YjdM was shown to catalyze phosphonoacetate hydrolase activity in vitro, but multiple lines of evidence suggest this is not its biological function: (1) The initial report linking YjdM to phosphonate catabolism was refuted by additional genetic analyses (Metcalf and Wanner 1993). (2) The experimentally validated PhnA belongs to a nonhomologous family, and expression of that family in E. coli showed PhnA activity was absent. (3) Genome neighborhood analysis shows yjdM genes are generally NOT near phosphonate catabolism or transport genes except in E. coli, where it happens to be upstream of the phn operon. (4) True phnA genes cluster with phosphonoacetate transporters and regulators. The in vitro activity may represent promiscuous hydrolytic activity rather than the evolved biological function.
Supporting Evidence:
PMID:40703034: "except for E. coli, yjdM genes are generally not close to phosphonate catabolism or transport genes"
PMID:40703034: "the PhnA activity observed in vitro is not supported, and additional in vivo experiments are required to confirm the biological role of this enzyme. This prediction was given a CS of 1"