Protein Thermostability Research Articles

High-throughput identification of calcium-regulated proteins across diverse proteomes

Calcium ions play important roles in nearly every biological process, yet whole-proteome analysis of calcium effectors has been hindered by a lack of high-throughput, unbiased, and quantitative methods to identify protein-calcium engagement. To address this, we adapted protein thermostability assays in budding yeast, human cells, and mouse mitochondria. Based on calcium-dependent thermostability, we identified 2,884 putative calcium-regulated proteins across human, mouse, and yeast proteomes. These data revealed calcium engagement of signaling hubs and cellular processes, including metabolic enzymes and the spliceosome. Cross-species comparison of calcium-protein engagement and mutagenesis experiments identified residue-specific cation engagement, even within well-known EF-hand domains. Additionally, we found that the dienoyl-coenzyme A (CoA) reductase DECR1 binds calcium at physiologically relevant concentrations with substrate-specific affinity, suggesting direct calcium regulation of mitochondrial fatty acid oxidation. These discovery-based proteomic analyses of calcium effectors establish a key resource to dissect cation engagement and its mechanistic effects across multiple species and diverse biological processes.

QProtein: Exploring Physical Features of Protein Thermostability Based on Structural Proteomics.

Thermostability, which is essential for the functional performance of enzymes, is largely determined by intramolecular physical interactions. Although many tools have been developed, existing computational methods have struggled to find the universal principles of protein thermostability. Recent advancements in structural proteomics have been driven by the introduction of deep neural networks such as AlphaFold2 and ESMFold. These innovations have enabled the characterization of protein structures with unprecedented speed and accuracy. Here, we introduce qProtein, a Python-implemented workflow designed for the quantitative analysis of physical interactions on the scale of structural proteomics. This platform accepts protein sequences as input and produces four structural features, including hydrophobic clusters, hydrogen bonds, electrostatic interactions, and disulfide bonds. To demonstrate the use of qProtein, we investigate the structural features related to protein thermostability in six glycoside hydrolase (GH) families, comprising a total of 3,811 protein structures. Our results indicate that in five enzyme families (GH11, GH12, GH5_2, GH10, and GH48), the thermophilic enzymes have a larger average area of hydrophobic clusters compared to the nonthermophilic enzymes within each family. Furthermore, our analysis of the local-structure regions reveals that the hydrophobic clusters are predominantly distributed in the distal regions of the GH11 enzymes. In addition, the average hydrophobic cluster area of the thermophilic enzymes is significantly higher than that of the nonthermophilic enzymes in the distal regions of the GH11 enzymes. Therefore, qProtein is a well-suited platform for analyzing the structural features of thermal stability at the level of structural proteomics. We provide the source code for qProtein at https://github.com/bj600800/qProtein, and the web server is available at http://qProtein.sdu.edu.cn:8888.

More is not always better. Multiplicity of Carbohydrate-binding domains for efficient starch hydrolysis activity

Lactobacillus amylovorus α-amylase is an endoenzyme with a peculiar starch-binding domain containing five identical family 26 carbohydrate-binding modules (CBM26). To investigate the impact of CBMs on catalytic activity, C-terminally truncated derivatives were constructed. The catalytic domain alone shows low affinity for the substrate and a very slow reaction rate, highlighting the importance of CBMs in maintaining the enzyme's optimal conformation and dynamics for efficient catalysis. CBMs enhance enzyme performance, as indicated by improved catalytic efficiency (kcat/Km). Interestingly, the enzyme variant LaCD3CBM, with three CBMs, exhibits the best catalytic efficiency on soluble starch, outperforming the wild-type amylase, with five CBMs. In the case of insoluble starch, the catalytic domain alone could not hydrolyze it and even adding a CBM, the release of reducing sugars was very inefficient. However, this efficiency was significantly improved by two orders of magnitude for the three-CBM variant and the wild-type amylase. CBMs also played a crucial role in protein thermostability, contributing to a higher melting temperature of the catalytic domain with just a single CBM. Notably, thermostability did not increase with the number of CBMs. In conclusion, spatial arrangement and interactions between the catalytic domain and CBMs significantly influenced enzymatic efficiency with both soluble and insoluble substrates. These interactions optimize enzymatic activity and improve thermostability.

Thermostable Nucleoid Protein Cren7 Slides Along DNA and Rapidly Dissociates From DNA While Not Inhibiting the Sliding of Other DNA-binding Protein

A nucleoid protein Cren7 compacts DNA, contributing to the living of Crenarchaeum in high temperature environment. In this study, we investigated the dynamic behavior of Cren7 on DNA and its functional relation using single-molecule fluorescence microscopy. We found two mobility modes of Cren7, sliding along DNA and pausing on it, and the rapid dissociation kinetics from DNA. The salt dependence analysis suggests a sliding with continuous contact to DNA, rather than hopping/jumping. The mutational analysis demonstrates that Cren7 slides along DNA while Trp (W26) residue interacts with the DNA. Furthermore, Cren7 does not impede the target search by a model transcription factor p53, implying no significant interference to other DNA-binding proteins on DNA. At high concentration of Cren7, the molecules form large clusters on DNA via bridging, which compacts DNA. We discuss how the dynamic behavior of Cren7 on DNA enables DNA-compaction and protein-bypass functions.

A High-Throughput Screening Platform for Engineering Poly(ethylene Terephthalate) Hydrolases.

The ability of enzymes to hydrolyze the ubiquitous polyester, poly(ethylene terephthalate) (PET), has enabled the potential for bioindustrial recycling of this waste plastic. To date, many of these PET hydrolases have been engineered for improved catalytic activity and stability, but current screening methods have limitations in screening large libraries, including under high-temperature conditions. Here, we developed a platform that can simultaneously interrogate PET hydrolase libraries of 104-105 variants (per round) for protein solubility, thermostability, and activity via paired, plate-based split green fluorescent protein and model substrate screens. We then applied this platform to improve the performance of a benchmark PET hydrolase, leaf-branch compost cutinase, by directed evolution. Our engineered enzyme exhibited higher catalytic activity relative to the benchmark, LCC-ICCG, on amorphous PET film coupon substrates (∼9.4% crystallinity) in pH-controlled bioreactors at both 65 °C (8.5% higher conversion at 48 h and 38% higher maximum rate, at 2.9% substrate loading) and 68 °C (11.2% higher conversion at 48 h and 43% higher maximum rate, at 16.5% substrate loading), up to 48 h, highlighting the potential of this screening platform to accelerate enzyme development for PET recycling.

A Thermostable Heme Protein Fold Adapted for Stereoselective C-H Bond Hydroxylation Using Peroxygenase Activity.

Thermostable protein folds of natural and synthetic origin are highly sought-after templates for biocatalyst generation due to their enhanced stability to elevated temperatures which overcomes one of the major limitations of applying enzymes for synthesis. Cytochrome P450 enzymes (CYPs) are a family of heme-thiolate monooxygenases that catalyse the oxidation of their substrates in a highly stereo- and regio-selective manner. The CYP enzyme (CYP107PQ1) from the thermophilic bacterium Meiothermus ruber binds the norisoprenoid β-ionone and was employed as a scaffold for catalyst design. The I-helix was modified to convert this enzyme from a monooxygenase into a peroxygenase (CYP107PQ1QE), enabling the enantioselective oxidation of β-ionone to (S)-4-hydroxy-β-ionone (94% e.e.). The enzyme was resistant to 20 mM H2O2, 20% (v/v) of organic solvent, supported over 1700 turnovers and was fully functional after incubation at 60 °C for 1 h and 30 °C for 365 days. The reaction was scaled-up to generate multi milligram quantities of the product for characterisation. Overall, we demonstrate that sourcing a CYP protein fold from an extremophile enabled the design of a highly stable enzyme for stereoselective C-H bond activation only using H2O2 as the oxidant, providing a viable strategy for future biocatalyst design.

The use of thermostable fluorescent proteins for live imaging in Sulfolobus acidocaldarius.

Among hyperthermophilic organisms, in vivo protein localization is challenging due to the high growth temperatures that can disrupt proper folding and function of mostly mesophilic-derived fluorescent proteins. While protein localization in the thermophilic model archaeon S. acidocaldarius has been achieved using antibodies with fluorescent probes in fixed cells, the use of thermostable fluorescent proteins for live imaging in thermophilic archaea has so far been unsuccessful. Given the significance of live protein localization in the field of archaeal cell biology, we aimed to identify fluorescent proteins for use in S. acidocaldarius. We expressed various previously published and optimized thermostable fluorescent proteins along with fusion proteins of interest and analyzed the cells using flow cytometry and (thermo-) fluorescent microscopy. Of the tested proteins, thermal green protein (TGP) exhibited the brightest fluorescence when expressed in Sulfolobus cells. By optimizing the linker between TGP and a protein of interest, we could additionally successfully fuse proteins with minimal loss of fluorescence. TGP-CdvB and TGP-PCNA1 fusions displayed localization patterns consistent with previous immunolocalization experiments. These initial results in live protein localization in S. acidocaldarius at high temperatures, combined with recent advancements in thermomicroscopy, open new avenues in the field of archaeal cell biology. This progress finally enables localization experiments in thermophilic archaea, which have so far been limited to mesophilic organisms.

Accurate Identification of Periplasmic Urea-binding Proteins by Structure- and Genome Context-assisted Functional Analysis

ABC transporters are ancient and ubiquitous nutrient transport systems in bacteria and play a central role in defining lifestyles. Periplasmic solute-binding proteins (SBPs) are components that deliver ligands to their translocation machinery. SBPs have diversified to bind a wide range of ligands with high specificity and affinity. However, accurate assignment of cognate ligands remains a challenging problem in SBPs. Urea metabolism plays an important role in the nitrogen cycle; anthropogenic sources account for more than half of global nitrogen fertilizer. We report identification of urea-binding proteins within a large SBP sequence family that encodes diverse functions. By combining genetic linkage between SBPs, ABC transporter components, enzymes or transcription factors, we accurately identified cognate ligands, as we verified experimentally by biophysical characterization of ligand binding and crystallographic determination of the urea complex of a thermostable urea-binding homolog. Using three-dimensional structure information, these functional assignments were extrapolated to other members in the sequence family lacking genetic linkage information, which revealed that only a fraction bind urea. Using the same combined approaches, we also inferred that other family members bind various short-chain amides, aliphatic amino acids (leucine, isoleucine, valine), γ-aminobutyrate, and as yet unknown ligands. Comparative structural analysis revealed structural adaptations that encode diversification in these SBPs. Systematic assignment of ligands to SBP sequence families is key to understanding bacterial lifestyles, and also provides a rich source of biosensors for clinical and environmental analysis, such as the thermostable urea-binding protein identified here.

Molecular analysis of changes in DNA binding affinity and bending extent induced by the mutations on the aromatic amino residues in Cren7.

Proteins from Crenarchaeal organisms exhibit remarkable thermal stability. The aromatic amino acids in Cren7, a Crenarchaeal protein, regulate protein stability and further modulate DNA binding and its compaction. Specific aromatic amino acids were mutated, and using spectroscopic and theoretical approaches, we have examined the structure, DNA binding affinity, and DNA bending ability of mutants. and compared with wild-type (WT) Cren7. The reverse titration profiles were analysed by a noncooperativeMcGhee-von Hippel model to estimate affinity constant (Ka) and site size (n) associated with binding to the DNA. Biolayer interferometry (BLI) measurements showed that the binding affinity decreased at higher salt concentrations. For theoretical analysis of extent of DNA bending, radius of gyration and bending angle were compared for WT and mutants. Time evolution of order parameters based on translational and rotational motion of tryptophan residue (W26) was used for qualitative detection of stacking interactions between W26 of Cren7 and DNA nucleobases. It was observed that orientation of W26 in F41A favoredformation of a new lone pair-lone pair interaction between DNA and Cren7. Consequently, in thermostable proteins, the aromatic residues at the terminus maintain structural stability, whereas the residues at the core optimize the degree of DNA bending and compaction.

ThermoLink: Bridging disulfide bonds and enzyme thermostability through database construction and machine learning prediction.

Disulfide bonds, covalently formed by sulfur atoms in cysteine residues, play a crucial role in protein folding and structure stability. Considering their significance, artificial disulfide bonds are often introduced to enhance protein thermostability. Although an increasing number of tools can assist with this task, significant amounts of time and resources are often wasted owing to inadequate consideration. To enhance the accuracy and efficiency of designing disulfide bonds for protein thermostability improvement, we initially collected disulfide bond and protein thermostability data from extensive literature sources. Thereafter, we extracted various sequence- and structure-based features and constructed machine-learning models to predict whether disulfide bonds can improve protein thermostability. Among all models, the neighborhood context model based on the Adaboost-DT algorithm performed the best, yielding "area under the receiver operating characteristic curve" and accuracy scores of 0.773 and 0.714, respectively. Furthermore, we also found AlphaFold2 to exhibit high superiority in predicting disulfide bonds, and to some extent, the coevolutionary relationship between residue pairs potentially guided artificial disulfide bond design. Moreover, several mutants of imine reductase 89 (IR89) with artificially designed thermostable disulfide bonds were experimentally proven to be considerably efficient for substrate catalysis. The SS-bond data have been integrated into an online server, namely, ThermoLink, available at guolab.mpu.edu.mo/thermoLink.

The Crystal Structure of Thermal Green Protein Q66E (TGP-E) and Yellow Thermostable Protein (YTP-E) E148D

Thermal green protein Q66E (TGP-E) has previously shown increased thermal stability compared to thermal green protein (TGP), a thermal stable fluorescent protein produced through consensus and surface protein engineering. In this paper, we describe the protein crystal structure of TGP-E to 2.0 Å. This structure reveals alterations in the hydrogen bond network near the chromophore that may result in the observed increase in thermal stability. We compare the very stable TGP-E protein to the structure of a yellow mutant version of this protein YTP-E E148D. The structure of this mutant protein reveals the rationale for the observed low quantum yield and directions for future protein engineering efforts.

Reinforcing thermostability and pH robustness of exo-inulinase facilitated by ReverseTag/ReverseCatcher tagging system

Enhancing protein stability is pivotal in the field of protein engineering. Protein self-cyclization using peptide a tagging system has emerged as an effective strategy for augmenting the thermostability of target proteins. In this study, we utilized a novel peptide tagging system, ReverseTag/ReverseCatcher, which leverages intramolecular ester bond formation. Initially, we employed GFP as a model to validate the feasibility of cyclization mediated by ReverseTag/ReverseCatcher in improving the protein thermostability. Cyclized GFP (cGFP) retained 30 % of its relative fluorescence after a 30-min incubation at 100 °C, while both GFP and linear GFP (lGFP) completely lost their fluorescence within 5 min. Additionally, we applied this method to exo-inulinase (EXINU), resulting in a variant named cyclized EXINU (cEXINU). The T50 and t1/2 values of cEXINU exhibited significant enhancements of 10 °C and 10 min, respectively, compared to EXINU. Furthermore, post-cyclization, EXINU demonstrated a broad operational pH range from 5 to 10 with sustained catalytic activity, and cEXINU maintained a half-life of 960 min at pH 5 and 9. Molecular dynamics simulations were conducted to elucidate the mechanisms underlying the enhanced thermostability and pH robustness of EXINU following cyclization. This study highlights that cyclization substanitially enhances the stability of both highly stable protein GFP and low-stable protein EXINU, mediated by the ReverseTag/ReverseCatcher tagging system. The ReverseTag/ReverseCatcher tagging system proves to be a potent conjugation method, with potential applications in improving thermostability, pH robustness, and other areas of protein engineering.

Rational design of disulfide bonds to increase thermostability of Rhodococcus opacus catechol 1,2 dioxygenase.

Catechol 1,2 dioxygenase is a versatile enzyme with several potential applications. However, due to its low thermostability, its industrial potential is not being met. In this study, the thermostability of a mesophilic catechol 1,2 dioxygenase from the species Rhodococcus opacus was enhanced via the introduction of disulphide bonds into its structure. Engineered designs (56) were obtained using computational prediction applications, with a set of hypothesized selection criteria narrowing the list to 9. Following recombinant production and purification, several of the designs demonstrated substantially improved protein thermostability. Notably, variant K96C-D278C yielded improvements including a 4.6°C increase in T50, a 725% increase in half-life, a 5.5°C increase in Tm, and a >10-fold increase in total turnover number compared to wild type. Stacking of best designs was not productive. Overall, current state-of-the-art prediction algorithms were effective for design of disulfide-thermostabilized catechol 1,2 dioxygenase.

PON-Tm: A Sequence-Based Method for Prediction of Missense Mutation Effects on Protein Thermal Stability Changes.

Proteins, as crucial macromolecules performing diverse biological roles, are central to numerous biological processes. The ability to predict changes in protein thermal stability due to mutations is vital for both biomedical research and industrial applications. However, existing experimental methods are often costly and labor-intensive, while structure-based prediction methods demand significant computational resources. In this study, we introduce PON-Tm, a novel sequence-based method for predicting mutation-induced thermal stability variations in proteins. PON-Tm not only incorporates features predicted by a protein language model from protein sequences but also considers environmental factors such as pH and the thermostability of the wild-type protein. To evaluate the effectiveness of PON-Tm, we compared its performance to four well-established methods, and PON-Tm exhibited superior predictive capabilities. Furthermore, to facilitate easy access and utilization, we have developed a web server.

Computer‐Aided Protein Surface Modification Strategy to Improve the Thermostability of α‐Amylase

AbstractStability under high‐temperature environments is crucial for amylase to function in starch‐based industries. This study develops a method of modifying protein surfaces by combining computer‐aided tools including FoldX, PoPMuSiC, Discovery Studio, and I‐Mutant 2.0 with conserved sequence analysis. A truncated α‐amylase ∆AmyPTG from Parageobacillus thermoglucosidasius DSMZ 2542 is rationally designed through this method to improve its thermostability. Seven single‐site variants are constructed and five of them displayed enhanced thermostability. Next, three double‐site variants are constructed with one particularly successful double‐site variant N31RT213R, exhibiting a 4.3‐fold longer half‐life at 80 °C. Notably, the specific activity of N31RT213R reaches 10 567.16 U mg−1, higher than ∆AmyPTG (6645.43 U mg−1). When applied to the corn starch liquefaction reaction, the mutant N31RT213R gets a higher yield of product concentration of about 255.70 µg mL−1, compared to 190.72 µg mL−1 for ∆AmyPTG. Intramolecular forces analysis and surface electrostatic charges analysis are conducted to determine possible causes for the improvement. Also, molecular dynamics simulation is used to analyze the flexibility shifts of the entire protein. This innovative rational engineering approach has proven to be a successful strategy for the selection of hot spots for protein thermostability evolution and has the potential to be applied to other enzymes.

Formulation Development of a COVID-19 Recombinant Spike Protein-Based Vaccine.

The purpose of this study was to develop a formulation for a recombinant prefusion spike protein vaccine against SARS-CoV-2. It was found that the spike protein was susceptible to aggregation due to mechanical stress. Therefore, formulation studies were initiated focused on screening pharmaceutical excipients capable of preventing this. The screening of a panel of potential stabilizing conditions found that Tween 20 could inhibit mechanically induced aggregation. A concentration-dependent study indicated that a higher concentration of Tween 20 (0.2% v/v) was required to prevent conformational changes in the trimer. The conformational changes induced by mechanical stress were characterized by size exclusion chromatography (SEC) and hydrogen-deuterium exchange mass spectrometry (HDX-MS), indicating the formation of an extended trimeric conformation that was also unable to bind to antibodies directed to the S2 domain. Long-term stability modeling, using advanced kinetic analysis, indicated that the formulation containing 0.2% (v/v) Tween 20 at a neutral pH was predicted to be stable for at least two years at 2 °C to 8 °C. Additional stabilizer screening conducted by thermal shift assay indicated that sucrose and glycerol were able to significantly increase the spike protein melting temperature (Tm) and improve the overall thermostability of the spike protein in a short-term stability study. Thus, while 0.2% (v/v) Tween 20 was sufficient to prevent aggregation and to maintain spike protein stability under refrigeration, the addition of sucrose further improved vaccine thermostability. Altogether, our study provides a systematic approach to the formulation of protein-based COVID-19 vaccine and highlights the impact of mechanical stress on the conformation of the spike protein and the significance of surfactants and stabilizers in maintaining the structural and functional integrity of the spike protein.

Molecular insights into β-Galactosidase enzyme for its potential application in food industry

β-galactosidase generally catalyzes the hydrolysis of β-d-galactosides and α-L-arabinosides. β-galactosidases are well known for its wide range of applications in food industry. In the present study, β-galactosidase from many sources, including archaea, bacteria, fungi, plants, and animals, were analyzed using computational studies such as structural, functional analyses. Sixty sequences were selected for multiple sequence alignment, phylogenetic study, motif and Pfam recognition, identification of physiochemical characteristics. The sequences of β-galactosidase were grouped independently for phylogenetic analysis based on their source. Thermostable proteins were identified through physicochemical analysis. The selected thermostable proteins were then subjected to functional analysis, transmembrane region prediction, secondary structure and domain architecture analysis. The selected highly thermostable sequences were rich in hydrophobic amino acids. Most of the selected β-galactosidase sequences are composed of random coils. The 3D structure of high thermostable bacterial β-galactosidase from Cryobacteriummelibiosiphilum (RJT89331.1) was predicted and the validation results obtained prove that the built structure is of good quality. The modeled structure showed highest similarity with crystal structure of Bifidobacterium longum β-galactosidase. Finally, β-galactosidase from Cryobacterium melibiosiphilum (RJT89331.1) was subjected to docking studies using the CB-DOCK tool with lactose, galactose, galactan, and pectin, with and their binding affinities are −6.6, −5.4, −6.0, −6.4 kJ/mol, respectively. The molecular dynamics simulation was conducted for the docked complex that exhibited the lowest binding energy using NMA (normal mode analysis) to observe the stability and motion of the complex. This computational investigation suggests that the β-galactosidase protein could help in screening β-galactosidase from suitable source for food industry applications.

Development of a New Non-Invasive Approach to Treat of Infantile Hemangioma with Growth-Inhibiting Endogenous Factors (Non-Invasive Approach to Treat of Infantile Hemangioma with Endogenous Factors)

Background: Infant hemangioma is a benign vascular tumor, the main reason for the development of which is genetic disorders occurring in embryogenesis. It should be noted that all methods of treating the disease, both general and local, have potential negative risks and side effects. The identification of endogenous regulatory factors and researching the possibilities of using them for treatment is relevant. There is the thermostable proteins complex (TPC) which has been identified in the cells of phylogenetically distant almost all organisms, which has the ability to decrease the mitotic activity of cells by inhibition of transcription. In the experimental model of hemangioma, it has been established that inhibition of capillary growth can only be achieved by injection. The aim of the work was to develop a new, less invasive method of hemangioma treatment with endogenous factors.

Thermostable α-Amylases and Laccases: Paving the Way for Sustainable Industrial Applications

The growing demand in industrial and biotechnological settings for more efficient enzymes with enhanced biochemical features, particularly thermostability and thermotolerance, necessitates a timely response. Renowned for their versatility, thermostable enzymes offer significant promise across a range of applications, including agricultural, medicinal, and biotechnological domains. This comprehensive review summarizes the structural attributes, catalytic mechanisms, and connection between structural configuration and functional activity of two major classes of thermostable enzymes: α-amylases and laccases. These enzymes serve as valuable models for understanding the structural foundation behind the thermostability of proteins. By highlighting the commercial importance of thermostable enzymes and the interest these generate among researchers in further optimization and innovation, this article can greatly contribute to ongoing research on thermostable enzymes and aiding industries in optimizing production processes via immobilization, use of stabilizing additives, chemical modification, protein engineering (directed evolution and mutagenesis), and genetic engineering (through cloning and expression of thermostable genes). It also gives insights to the exploration of suitable strategies and factors for enhancing thermostability like increasing substrate affinity; introducing electrostatic, intramolecular, and intermolecular hydrophobic interactions; mitigating steric hindrance; increasing flexibility of an active site; and N- and C-terminal engineering, thus resulting in heightened multipronged stability and notable enhancements in the enzymes’ industrial applicability.

Progress in the application of artificial intelligence-assisted molecular modification of enzymes

Natural enzymes are often difficult to meet the needs of application and research in terms of activity, enantiomer selectivity or thermal stability. Therefore, it is an important task of enzyme engineering to explore efficient molecular modification technologies to improve the properties of such enzymes. The molecular modification technologies of enzymes mainly include rational design, directed evolution, and artificial intelligence-assisted design. Directed evolution and rational design are experiment-driven molecular modification approaches of enzymes and have been successfully applied to enzyme engineering. However, due to the huge space sizes of protein sequences and the lack of experimental data, the current modification methods still face major challenges. With the development of next-generation sequencing, high-throughput screening, protein databases, and artificial intelligence (AI), data-driven enzyme engineering is emerging as a promising solution to these challenges. The AI-assisted statistical learning method has been used to establish a model for predicting the sequence/structure-properties of enzymes in a data-driven manner. Excellent mutant enzymes can be selected according to the prediction results, which greatly improve the efficiency of molecular modification. Considering the application requirements of molecular modification of enzymes, this paper reviews the data acquisition methods and application examples of AI-assisted molecular modification of enzymes, with focuses on the convolutional neural network method for predicting protein thermostability, aiming to provide reference for researchers in this field.