Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 111 of the 112 references it cites were found in a registry or an index; 107 of the 144 sources its claims cite were checked for support.

Toward the Complete Functional Characterization of a Minimal Bacterial Proteome

Open access

Abstract

Recently, we presented a whole-cell kinetic model of the genetically minimal bacterium JCVI-syn3A that described the coupled metabolic and genetic information processes and predicted behaviors emerging from the interactions among these networks. JCVI-syn3A is a genetically reduced bacterial cell that has the fewest number and smallest fraction of genes of unclear function, with approximately 90 of its 452 protein-coding genes (that is less than 20%) unannotated. Further characterization of unclear JCVI-syn3A genes strengthens the robustness and predictive power of cell modeling efforts and can lead to a deeper understanding of biophysical processes and pathways at the cell scale. Here, we apply computational analyses to elucidate the functions of the products of several essential but previously uncharacterized genes involved in integral cellular processes, particularly those directly affecting cell growth, division, and morphology. We also suggest directed wet-lab experiments informed by our analyses to further understand these "missing puzzle pieces" that are an essential part of the mosaic of biological interactions present in JCVI-syn3A. Our workflow leverages evolutionary sequence analysis, protein structure prediction, interactomics, and genome architecture to determine upgraded annotations. Additionally, we apply the structure prediction analysis component of our work to all 452 protein coding genes in JCVI-syn3A to expedite future functional annotation studies as well as the inverse mapping of the cell state to more physical models requiring all-atom or coarse-grained representations for all JCVI-syn3A proteins.

Overview

CoChat Summary

TL;DR

The paper applies AlphaFold2-based structure prediction, sequence analysis, interactomics, and genome-locality analysis to annotate previously unclear genes in the minimal bacterium JCVI-syn3A. It upgrades annotations for 25 of the approximately 90 unclear protein-coding genes and provides structural models for all 452 JCVI-syn3A proteins.

Key findings

  • JCVI-syn3A has 493 genes, 543 kbp DNA, and 452 protein-coding genes; approximately 90 protein-coding genes (less than 20%) were previously unannotated. The authors upgraded annotations for 25 of these genes and provided additional information for tens of the remaining protein-coding genes.
  • The paper characterizes 12 genes in Table 1, including ezrA/0239 as an FtsZ filament regulator, gpsB/0353 as a divisome localization regulator, secDF/0412 as a protein translocase subunit, yqkD/0520 as a membrane lipase/aminopeptidase, ylbN/0527 as a 23S rRNA regulator, and atpI/0797 as an ATP synthase component.
  • Structural alignments supported assignments with significant FATCAT p-values, including gpsB/0353 (p = 2.72 × 10–6, RMSD 1.10 Å, 100% aligned), ezrA/0239 (p = 3.01 × 10–10, RMSD 4.64 Å, 84% aligned), yqkD/0520 (p < 1.0 × 10–20, RMSD 3.00 Å, 96% aligned), and atpI/0797 (p = 1.81 × 10–11, RMSD 1.22 Å, 92% aligned).
  • The authors suggest directed wet-lab experiments to validate several annotations, including proteomic/transcriptomic tests for ylbN/0527, membrane protease assays for yqkD/0520, GTPase assays for ezrA/0239, and colocalization for atpI/0797 and ylxM/0430.

Methods

  • Computational workflow combining AlphaFold2 and RoseTTAFold structure prediction, sequence alignment (CLUSTALW, HMMER, HHPred), protein threading, pairwise structural alignment (FATCAT, STAMP), interactomics (SynWiki/STRING), and genome-locality analysis.
  • AlphaFold2 predictions used default CASP14-like parameters: 3 recycles, 5 models, full 3+ terabyte databases, and PDB entries deposited on or before 9-15-2021; structures were also relaxed via AMBER, and predictions were generated for all 452 protein-coding genes.
  • Functional assignments were made when predicted structures aligned significantly with experimental or predicted structures from related bacteria, with additional support from sequence motifs, interactomes, and conserved genome architecture.

Limitations

  • The paper states that phenotypes of many annotated genes are difficult to study because the genes are essential and deletion strains cannot be obtained.
  • The paper states that moonlighting proteins in mycoplasmas can confound function prediction, and that low AlphaFold2 prediction confidence can occur especially in and around membrane-embedded domains.
  • The paper states that some assignments remain speculative, such as the proposed linkage between ylbN/0527 removal and irregular daughter-cell morphology, which requires further experimental study; more broadly, the annotations are computational predictions and the text makes evident that they are not experimentally validated in this work.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1021/acs.jpcb.2c04188

Cites

111 of 111 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 (A) Predicted structures of gpsB/0353 from AlphaFold2 (red) and RoseTTAFold (blue) aligned with the experimental structure of gpsB from B. subtilis (RCSB PDB: 4UG3)58 at the N-terminal coiled-coil domain using STAMP.41 (B) Predicted structure of lemA/0604 (see section 3) from RoseTTAFold (blue) aligned by FATCAT25 with the lemA two-component cell growth regulator from T. maritima (yellow, RCSD PDB: 2ETD(61)). 142 of the 210 residues (68%) are well-aligned with a RMSD of only 1.67 Å. (C) Multiple sequence alignment of /0353 from JCVI-syn3A with sequences of gpsB from B. subtilis (strain 168) and S. aureus (strain NCTC 8325/PS 47) shows the conserved C-terminal domain consensus sequence,59 lending further evidence to the assignment of gpsB/0353 instead of its paralogue divIVA. All images visualized in VMD.42
Figure 1 (A) Predicted structures of gpsB/0353 from AlphaFold2 (red) and RoseTTAFold (blue) aligned with the experimental structure of gpsB from B. subtilis (RCSB PDB: 4UG3)58 at the N-terminal coiled-coil domain using STAMP.41 (B) Predicted structure of lemA/0604 (see section 3) from RoseTTAFold (blue) aligned by FATCAT25 with the lemA two-component cell growth regulator from T. maritima (yellow, RCSD PDB: 2ETD(61)). 142 of the 210 residues (68%) are well-aligned with a RMSD of only 1.67 Å. (C) Multiple sequence alignment of /0353 from JCVI-syn3A with sequences of gpsB from B. subtilis (strain 168) and S. aureus (strain NCTC 8325/PS 47) shows the conserved C-terminal domain consensus sequence,59 lending further evidence to the assignment of gpsB/0353 instead of its paralogue divIVA. All images visualized in VMD.42CC BY-NC-ND · PubMed Central
Figure 2 (A) Structural alignment by FATCAT25 of the predicted AlphaFold2 structure for gene /0239 (blue) to the structure of the cell division regulatory protein EzrA from B. subtilis (yellow, RCSB PDB: 4UXV(65)) with 498 or 84% of the residues of /0239 being well-aligned to the experimental structure with a p-value of 3.01 × 10–10 for significant structural similarity. (B) Hydrophobicity coloring (red, hydrophobic resiudes; blue, hydrophilic residues) for the AlphaFold2 predicted structure for /0239. Hydrophobic regions at the N and C termini match the previous findings that EzrA binds the membrane at each terminal domain.65 (C) Sequence Alignment of gene /0239 with the sequences for ezrA from S. aureus and B. subtilis with BLOSUM50 similarity score coloring. While the entire characteristic “QNR” patch63,64 is not present in JCVI-syn3A, significant sequence similarity remains and the nearly universally conserved asparagine residue located centrally in the patch is present. (D) The genome architecture of JCVI-syn3A also lends support to the assignment of ezrA/0239. Just as in the related Gram-positive organism B. subtilis, the gene coding for EzrA lies between the genes for thiL coding for thiamine monophosphate kinase and rpsD coding for ribosomal protein S2 on opposing strands of the circular genome.
Figure 2 (A) Structural alignment by FATCAT25 of the predicted AlphaFold2 structure for gene /0239 (blue) to the structure of the cell division regulatory protein EzrA from B. subtilis (yellow, RCSB PDB: 4UXV(65)) with 498 or 84% of the residues of /0239 being well-aligned to the experimental structure with a p-value of 3.01 × 10–10 for significant structural similarity. (B) Hydrophobicity coloring (red, hydrophobic resiudes; blue, hydrophilic residues) for the AlphaFold2 predicted structure for /0239. Hydrophobic regions at the N and C termini match the previous findings that EzrA binds the membrane at each terminal domain.65 (C) Sequence Alignment of gene /0239 with the sequences for ezrA from S. aureus and B. subtilis with BLOSUM50 similarity score coloring. While the entire characteristic “QNR” patch63,64 is not present in JCVI-syn3A, significant sequence similarity remains and the nearly universally conserved asparagine residue located centrally in the patch is present. (D) The genome architecture of JCVI-syn3A also lends support to the assignment of ezrA/0239. Just as in the related Gram-positive organism B. subtilis, the gene coding for EzrA lies between the genes for thiL coding for thiamine monophosphate kinase and rpsD coding for ribosomal protein S2 on opposing strands of the circular genome.CC BY-NC-ND · PubMed Central
Figure 3 (A) Structural alignment of the AlphaFold2 predicted structure of the YqkD esterase/lipase/serine aminopeptidase from B. subtilis,68 (yellow) to the predicted protein structure of /0520 from AlphaFold2 (blue) using FATCAT 2.0.25 The alignment characterizes the structures as being significantly similar to more than 95% of the residues of /0520 being well-aligned to the B. subtilis protein predicted structure. (B) Structural alignment via STAMP of P. furiosis PF2001 hydrolase (yellow, RCSB PDB: 5G59(69)) to the predicted structures for /0520 from both AlphaFold2 (red) and RoseTTAFold (blue). (C) The alignment with Qres coloring43 shows that the hydrolase helical bundle is highly conserved (blue-white-red color scheme of decreasing conservation), supporting the assignment of /0520 as an aminopeptidase/esterase/lipase, of the α–β hydrolase superfamily (see ref (66) and Pfam), and suggesting that PF2001 might play a similar role beyond its confirmed general esterase activity.
Figure 3 (A) Structural alignment of the AlphaFold2 predicted structure of the YqkD esterase/lipase/serine aminopeptidase from B. subtilis,68 (yellow) to the predicted protein structure of /0520 from AlphaFold2 (blue) using FATCAT 2.0.25 The alignment characterizes the structures as being significantly similar to more than 95% of the residues of /0520 being well-aligned to the B. subtilis protein predicted structure. (B) Structural alignment via STAMP of P. furiosis PF2001 hydrolase (yellow, RCSB PDB: 5G59(69)) to the predicted structures for /0520 from both AlphaFold2 (red) and RoseTTAFold (blue). (C) The alignment with Qres coloring43 shows that the hydrolase helical bundle is highly conserved (blue-white-red color scheme of decreasing conservation), supporting the assignment of /0520 as an aminopeptidase/esterase/lipase, of the α–β hydrolase superfamily (see ref (66) and Pfam), and suggesting that PF2001 might play a similar role beyond its confirmed general esterase activity.CC BY-NC-ND · PubMed Central
Figure 4 (A) Genome locality of ylbN and ylbN/0527 in B. subtilis and JCVI-syn3A, respectively. The ylbN genes are colored in red, while adjacent genes located prior to the dcw cluster in both organisms include rpmF, mraZ, and rsmH (also known as mraW) (blue), with additional genes only present in B. subtilis shown (green). Both gene clusters show similar orientation and arrangement on the reverse strand. (B) Structural alignment of AlphaFold2 prediction of ylbN/0527 (blue) and the predicted AlphaFold2 structure for ylbN in B. subtilis 168 (yellow76). (C) CLUSTALW sequence alignment of ylbN/0527 and ylbN from B. subtilis with BLOSUM50 similarity coloring. The conserved 158R reported in ref (75) is aligned in the “DPR” region of both sequences.
Figure 4 (A) Genome locality of ylbN and ylbN/0527 in B. subtilis and JCVI-syn3A, respectively. The ylbN genes are colored in red, while adjacent genes located prior to the dcw cluster in both organisms include rpmF, mraZ, and rsmH (also known as mraW) (blue), with additional genes only present in B. subtilis shown (green). Both gene clusters show similar orientation and arrangement on the reverse strand. (B) Structural alignment of AlphaFold2 prediction of ylbN/0527 (blue) and the predicted AlphaFold2 structure for ylbN in B. subtilis 168 (yellow76). (C) CLUSTALW sequence alignment of ylbN/0527 and ylbN from B. subtilis with BLOSUM50 similarity coloring. The conserved 158R reported in ref (75) is aligned in the “DPR” region of both sequences.CC BY-NC-ND · PubMed Central
Figure 5 (A) Genome architecture of the related Gram-positive bacterium B. subtilis near the gene coding for SecDF translocon components and that adjacent to gene/0412 show conservation. (B) STAMP structural alignment with visualization in VMD of the AlphaFold2 predicted structure of gene 0430/ylxM (red) and the experimentally determined structure for YlxM-like effector of the signal recognition particle from S. aureus (yellow, RCSD PDB: 1S7O(82)). (C) Qres coloring43 (with a blue-white-red color palette of decreasing structural similarity) shows a high degree of structural conservation between the predicted and experimental structures.
Figure 5 (A) Genome architecture of the related Gram-positive bacterium B. subtilis near the gene coding for SecDF translocon components and that adjacent to gene/0412 show conservation. (B) STAMP structural alignment with visualization in VMD of the AlphaFold2 predicted structure of gene 0430/ylxM (red) and the experimentally determined structure for YlxM-like effector of the signal recognition particle from S. aureus (yellow, RCSD PDB: 1S7O(82)). (C) Qres coloring43 (with a blue-white-red color palette of decreasing structural similarity) shows a high degree of structural conservation between the predicted and experimental structures.CC BY-NC-ND · PubMed Central
Figure 6 (A) FATCAT25 structural alignment of AlphaFold2 predicted structure of /0516 (blue) and gplG protease from H. influenzae (RCSB PDB: 2NR9(89)), where only the lower helical portion of the protease is available in the experimental structure. (B) FATCAT structural alignment of the predicted structure from (A) with the Alphafold2 predicted structure for the gplG paralogue protease yqgP from B. subtilis shows excellent agreement to both the upper β- sheet and the lower -a-helical regions of the JCVI-syn3A structure, with a p-value of 7.19 × 10–9 showing significant structural similarity.
Figure 6 (A) FATCAT25 structural alignment of AlphaFold2 predicted structure of /0516 (blue) and gplG protease from H. influenzae (RCSB PDB: 2NR9(89)), where only the lower helical portion of the protease is available in the experimental structure. (B) FATCAT structural alignment of the predicted structure from (A) with the Alphafold2 predicted structure for the gplG paralogue protease yqgP from B. subtilis shows excellent agreement to both the upper β- sheet and the lower -a-helical regions of the JCVI-syn3A structure, with a p-value of 7.19 × 10–9 showing significant structural similarity.CC BY-NC-ND · PubMed Central
Figure 7 (A) Structural alignment of AlphaFold2 prediction for gabP/0878 (blue) with the predicted structure of the B. subtilis GabP permease (yellow). (B) FATCAT structural alignment of the predicted structure of yloU/0421 from JCVI-syn3A (blue) with the AlphaFold2 predicted structure for B. subtilis YloU fatty acid regulatory protein (yellow). (C) The genomic context in which yloU is found, adjacent to the fatty acid processing genes fakA and plsX, is conserved for gene yloU/0421 in JCVI-syn3A and has been observed to be conserved previously in the Firmicutes.91 (D) FATCAT structural alignment of predicted structure for gene yidC/0908 and predicted structure for M. pneumoniae show significant structural similarity.
Figure 7 (A) Structural alignment of AlphaFold2 prediction for gabP/0878 (blue) with the predicted structure of the B. subtilis GabP permease (yellow). (B) FATCAT structural alignment of the predicted structure of yloU/0421 from JCVI-syn3A (blue) with the AlphaFold2 predicted structure for B. subtilis YloU fatty acid regulatory protein (yellow). (C) The genomic context in which yloU is found, adjacent to the fatty acid processing genes fakA and plsX, is conserved for gene yloU/0421 in JCVI-syn3A and has been observed to be conserved previously in the Firmicutes.91 (D) FATCAT structural alignment of predicted structure for gene yidC/0908 and predicted structure for M. pneumoniae show significant structural similarity.CC BY-NC-ND · PubMed Central
Figure 8 (A) STAMP structural alignment of predicted structure for B. subtilis atpI (yellow) with predicted protein structures for atpI/0797 from AlphaFold2 (red) and RoseTTAFold (blue). Images generated using VMD.42 Alignment of these two structures obtained a FATCAT25p-value of greater than 1.0 × 10–11, showing significant similarity with 92% of the JCVI-syn3A structure being well-aligned to that predicted for B. subtilis. (B) FATCAT alignment of the Spinicia oleracea atpI structure (RSCB PDB: 6FKI(101)) with the AlphaFold2 predicted structure of atpI/0797 with conservation of the central two α-helical regions especially. Alignment of these two structures obtained a FATCAT25 gave 130 residues or 99% of the JCVI-syn3A structure being well-aligned to the larger 247 residue cryo-EM structure for Spinicia oleracea. (C) Genomic organization from JCVI-syn3A and B. subtilis of the ATP synthase operon3,49 is identical lending further support to the assignment of atpI/0797 (red, bottom-right).
Figure 8 (A) STAMP structural alignment of predicted structure for B. subtilis atpI (yellow) with predicted protein structures for atpI/0797 from AlphaFold2 (red) and RoseTTAFold (blue). Images generated using VMD.42 Alignment of these two structures obtained a FATCAT25p-value of greater than 1.0 × 10–11, showing significant similarity with 92% of the JCVI-syn3A structure being well-aligned to that predicted for B. subtilis. (B) FATCAT alignment of the Spinicia oleracea atpI structure (RSCB PDB: 6FKI(101)) with the AlphaFold2 predicted structure of atpI/0797 with conservation of the central two α-helical regions especially. Alignment of these two structures obtained a FATCAT25 gave 130 residues or 99% of the JCVI-syn3A structure being well-aligned to the larger 247 residue cryo-EM structure for Spinicia oleracea. (C) Genomic organization from JCVI-syn3A and B. subtilis of the ATP synthase operon3,49 is identical lending further support to the assignment of atpI/0797 (red, bottom-right).CC BY-NC-ND · PubMed Central

Verification

Claims the paper makes

60 Valid · 19 Not valid · 12 Not checked · 31 Own results and interpretations

52 more claims · 26 with their cited sources checked

Create a free account to see all 122 claims and to verify the 8 references still open.

See all claims — free account
  • It may state two separate claims as one. 46
  • Probably not a substantial claim. 21
  • The quote only partly states this claim. 5

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

5 of 7 · the rest aren’t loaded

2 more findings

Create a free account to see every finding of this check.

See all findings — free account

Citations

Cites

Cited by

See all 24

Metadata

Authors
Bianchi, D. M., Pelletier, J. F., Hutchison, C. A., Glass, J. I., & Luthey-Schulten, Z.
Affiliations
  • University of Illinois Urbana-Champaign (US)
  • Centro Nacional de Biotecnología (ES)
  • J. Craig Venter Institute (US)
Corresponding author
Luthey-Schulten, Zaida
Venue
The Journal of Physical Chemistry B
Published
2022-09-01
Volume
126
Issue
36
Pages
6820-6834
ISSN
  • 1520-5207 (ISSN-L)
  • 1520-6106
OpenAlex
W4294043777
Cited by
22
References
112
Citations by year
2026: 3 · 2025: 4 · 2024: 8 · 2023: 7
FWCI
1.55
Type
journal article
Language
en
Open access
Locations
  • The Journal of Physical Chemistry B (publishedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
Funders
  • Division of Molecular and Cellular Biosciences (MCB-1818344, MCB-1840320)
  • Human Frontier Science Program
  • Division of Physics (PHY-1430124, PHY-2014027)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Microbial Metabolic Engineering and Bioproduction
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
Keywords
unannotated genes, functional annotation, protein structure prediction, interactomics, genome organization, cell growth and division, cell morphology, synthetic biology
MeSH terms
Bacteria, Proteome

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.