CoChat
Sign Up
Cited by 56Enzyme Catalysis and ImmobilizationEnzyme Production and CharacterizationProtein Structure and Dynamics

The DOI matches its Crossref record; 68 of the 68 references it cites were found in a registry or an index.

Harnessing protein language model for structure-based discovery of highly efficient and robust PET hydrolases

Open access
Wu, Banghao, Zhong, Bozitao, Zheng, Lirong, Huang, Runye, Jiang, Shifeng, Li, Mingchen, Hong, Liang, Tan, Pan·2025·Nature Communications
DOI 10.1038/s41467-025-61599-zPMC PMC12228687

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Plastic waste, particularly polyethylene terephthalate (PET), presents significant environmental challenges, driving extensive research into enzymatic biodegradation. However, existing PET hydrolases (PETases) are limited by narrow sequence diversity and suboptimal performance. This study introduces VenusMine, a protein discovery pipeline that integrates protein language models (PLMs) with a representation tree to identify PETases based on structural similarity using sequence information. Using the crystal structure of Is PETase as a template, VenusMine identifies and clusters target proteins. Candidates are further screened using PLM-based assessments of solubility and thermostability, leading to the selection of 34 proteins for biochemical validation. Results reveal that 14 candidates exhibit PET degradation activity across 30–60 °C. Notably, a PET hydrolase from Kibdelosporangium banguiense ( Kb PETase) demonstrates a melting temperature (T m ) 32 °C higher than Is PETase and exhibits the highest PET degradation activity within 30 – 65 °C among wild-type PETases. Kb PETase also surpasses FastPETase and LCC in catalytic efficiency. X-ray crystallography and molecular dynamics simulations show that Kb PETase possesses a conserved catalytic domain and enhanced intramolecular interactions, underpinning its improved functionality and thermostability. This work demonstrates a novel deep learning approach for discovering natural PETases with enhanced properties.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-025-61599-z

Cites

68 of 68 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 In silico structure-based PETase discovery pipeline and clustering results. a The pipeline for PETase discovery combines structure retrieval with FoldSeek, sequence retrieval with MMSeqs2, protein embeddings computation via ProstT5, and representation tree construction to identify functional and robust PETase. b The UMAP visualization of the PETase candidates identified by ProstT5. Green points indicate proteins with previously validated PET hydrolytic activity or annotated with corrected EC number 3.1.1.101, while red points represent those not annotated with the corresponding EC number. Pink points represent proteins identified using FoldSeek, and orange points correspond to proteins discovered through sequence-based searches. c The UMAP visualization of the PETase candidates colored by representation tree clusters. d The representation tree constructed using embeddings from ProstT5. Green bars represent proteins with previously validated PETase activity, blue bars present enzymes with the correct EC number, while red bars indicate enzymes not annotated with the correct EC number. Two clusters are selected for further screening: the first cluster (dark red) and the third cluster (orange), arranged in a counterclockwise direction.
Fig. 1 In silico structure-based PETase discovery pipeline and clustering results. a The pipeline for PETase discovery combines structure retrieval with FoldSeek, sequence retrieval with MMSeqs2, protein embeddings computation via ProstT5, and representation tree construction to identify functional and robust PETase. b The UMAP visualization of the PETase candidates identified by ProstT5. Green points indicate proteins with previously validated PET hydrolytic activity or annotated with corrected EC number 3.1.1.101, while red points represent those not annotated with the corresponding EC number. Pink points represent proteins identified using FoldSeek, and orange points correspond to proteins discovered through sequence-based searches. c The UMAP visualization of the PETase candidates colored by representation tree clusters. d The representation tree constructed using embeddings from ProstT5. Green bars represent proteins with previously validated PETase activity, blue bars present enzymes with the correct EC number, while red bars indicate enzymes not annotated with the correct EC number. Two clusters are selected for further screening: the first cluster (dark red) and the third cluster (orange), arranged in a counterclockwise direction.CC BY · PubMed Central
Fig. 2 Experimental validation and biochemical characterization of candidates. a Enzyme activity of 14 proteins showing effective ester bond hydrolysis, as measured using p-nitrophenyl butyrate (pNPB) as a substrate at 37 °C. Reactions were performed in triplicate; data are presented as mean values ± SD. b The Tm of the 14 proteins exhibiting ester bond cleavage activity. Reactions were performed in triplicate; data are presented as mean values ± SD. c Comparison of the degradation activity of candidate proteins on PET films. The reaction was conducted in 50 mM Glycine-NaOH (pH 9.0) at various temperatures (30, 40, 50, 55, and 60 °C) for 72 h. Reactions were performed in triplicate; data are presented as mean values ± SD.
Fig. 2 Experimental validation and biochemical characterization of candidates. a Enzyme activity of 14 proteins showing effective ester bond hydrolysis, as measured using p-nitrophenyl butyrate (pNPB) as a substrate at 37 °C. Reactions were performed in triplicate; data are presented as mean values ± SD. b The Tm of the 14 proteins exhibiting ester bond cleavage activity. Reactions were performed in triplicate; data are presented as mean values ± SD. c Comparison of the degradation activity of candidate proteins on PET films. The reaction was conducted in 50 mM Glycine-NaOH (pH 9.0) at various temperatures (30, 40, 50, 55, and 60 °C) for 72 h. Reactions were performed in triplicate; data are presented as mean values ± SD.CC BY · PubMed Central
Fig. 3 Activity and thermostability measurements of discovered PETases. a PET film degradation activity of the previously characterized PETases compared to KbPETase. The reaction was conducted in 50 mM Glycine-NaOH (pH 9.0) at various temperatures (30, 40, 50, 55, 60 and 65 °C) for 72 h. Reactions were performed in triplicate; data are presented as mean values ± SD. b Comparison of the Tm of KbPETase with other reported WT PETases using DSF. Reactions were performed in triplicate; data are presented as mean values ± SD. c Comparison of the enzymatic kinetics curves of KbPETase, LCC and FastPETase using pNPB as the substrate. Reactions were performed in triplicate; data are presented as mean values ± SD. d The depolymerization percentages of PET by KbPETase and FastPETase at 50 °C, as well as by LCC at 65 °C, were determined through UPLC analysis of the released products. Reactions were conducted in triplicate, and the results are presented as mean values ± SD.
Fig. 3 Activity and thermostability measurements of discovered PETases. a PET film degradation activity of the previously characterized PETases compared to KbPETase. The reaction was conducted in 50 mM Glycine-NaOH (pH 9.0) at various temperatures (30, 40, 50, 55, 60 and 65 °C) for 72 h. Reactions were performed in triplicate; data are presented as mean values ± SD. b Comparison of the Tm of KbPETase with other reported WT PETases using DSF. Reactions were performed in triplicate; data are presented as mean values ± SD. c Comparison of the enzymatic kinetics curves of KbPETase, LCC and FastPETase using pNPB as the substrate. Reactions were performed in triplicate; data are presented as mean values ± SD. d The depolymerization percentages of PET by KbPETase and FastPETase at 50 °C, as well as by LCC at 65 °C, were determined through UPLC analysis of the released products. Reactions were conducted in triplicate, and the results are presented as mean values ± SD.CC BY · PubMed Central
Fig. 4 Structure and sequence characteristics of APET enzymes and known PETases. a Crystal structure of KbPETase (green, PDB: 9IW9) and IsPETase (blue, PDB: 5XH3). The substrate structure (yellow) is derived from the IsPETase structure and structurally aligned with the KbPETase pocket. Pocket structure of (b) IsPETase and (c) KbPETase. d Phylogenetic tree of the 34 candidate proteins selected for experimental validation (blue and green) alongside previously reported PETases (orange). The APET candidates with confirmed activity are marked with green, otherwise in blue. (e, f) Sequence logo plots for the pocket residues. The height of each letter indicates the conservation score for each site, indexed by the residue position in KbPETase, with catalytic traid marked in red. g Structure-based alignment of pocket residues from APET (green) and known PETases (orange). Residues are annotated by their biochemical properties, with hierarchical clustering presented on the right. Residue indices correspond to their positions in KbPETase.
Fig. 4 Structure and sequence characteristics of APET enzymes and known PETases. a Crystal structure of KbPETase (green, PDB: 9IW9) and IsPETase (blue, PDB: 5XH3). The substrate structure (yellow) is derived from the IsPETase structure and structurally aligned with the KbPETase pocket. Pocket structure of (b) IsPETase and (c) KbPETase. d Phylogenetic tree of the 34 candidate proteins selected for experimental validation (blue and green) alongside previously reported PETases (orange). The APET candidates with confirmed activity are marked with green, otherwise in blue. (e, f) Sequence logo plots for the pocket residues. The height of each letter indicates the conservation score for each site, indexed by the residue position in KbPETase, with catalytic traid marked in red. g Structure-based alignment of pocket residues from APET (green) and known PETases (orange). Residues are annotated by their biochemical properties, with hierarchical clustering presented on the right. Residue indices correspond to their positions in KbPETase.CC BY · PubMed Central
Fig. 5 Molecular dynamics simulation analyses of KbPETase, IsPETase, and LCC. a–c The distribution of Rg for KbPETase (a), LCC (b) and IsPETase (c). d–f Residue-wise root mean square fluctuation (RMSF) mapped onto structures, with green (KbPETase), red (LCC) and blue (IsPETase) for low RMSF and yellow for high RMSF. Residues in catalytic pockets are presented with sticks. g–i non-covalent interactions, including numbers of (g) global hydrogen bonds, (h) catalytic pocket hydrogen bonds, and (i) global salt bridges. Data are presented as mean values ± SD.
Fig. 5 Molecular dynamics simulation analyses of KbPETase, IsPETase, and LCC. a–c The distribution of Rg for KbPETase (a), LCC (b) and IsPETase (c). d–f Residue-wise root mean square fluctuation (RMSF) mapped onto structures, with green (KbPETase), red (LCC) and blue (IsPETase) for low RMSF and yellow for high RMSF. Residues in catalytic pockets are presented with sticks. g–i non-covalent interactions, including numbers of (g) global hydrogen bonds, (h) catalytic pocket hydrogen bonds, and (i) global salt bridges. Data are presented as mean values ± SD.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Cited by

Sorted by citations

See all 56

Related

See all

Metadata

Authors
Wu, B., Zhong, B., Zheng, L., Huang, R., Jiang, S., Li, M., Hong, L., & Tan, P.
Affiliations
Shanghai Jiao Tong University (CN)
Corresponding author
  • Zheng, Lirong
  • Hong, Liang
  • Tan, Pan
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2025-07-05
Volume
16
Issue
1
Pages
6211-6211
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-025-61599-z
PMC
PMC12228687
OpenAlex
W4412048048
Cited by
50
References
67
Citations by year
2026: 39 · 2025: 11
FWCI
14.93
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • Research Square (acceptedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Natural Science Foundation of China (12104295)
  • Shanghai Municipal Education Commission
  • Science and Technology Commission of Shanghai Municipality
  • Shanghai Jiao Tong University (21X010200843)
  • Centre Scientifique et Technique du Bâtiment
  • National Key Research and Development Program of China
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Enzyme Catalysis and Immobilization
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Biotechnology › Enzyme Production and Characterization
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
Keywords
PET hydrolase, polyethylene terephthalate degradation, plastic waste biodegradation, protein language models, thermal stability, catalytic efficiency, protein solubility prediction
MeSH terms
Hydrolases, Polyethylene Terephthalates, Protein Conformation, Crystallography, X-Ray, Molecular Dynamics Simulation

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.

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