The relationship between pH and enzyme catalytic activity, especially the optimal pH (pHopt) at which enzymes function, is critical for biotechnological applications. Hence, computational methods to predict pHopt will enhance enzyme discovery and design by facilitating accurate identification of enzymes that function optimally at specific pH levels, and by elucidating sequence–function relationships. Here we proposed and evaluated various machine learning methods for predicting pHopt, conducting extensive hyperparameter optimization and training over 11,000 model instances. Our results demonstrate that models utilizing language model embeddings markedly outperform other methods in predicting pHopt. We present EpHod, the best-performing model, to predict pHopt, making it publicly available to researchers. From sequence data, EpHod directly learns structural and biophysical features that relate to pHopt, including proximity of residues to the catalytic centre and the accessibility of solvent molecules. Overall, EpHod presents a promising advancement in pHopt prediction and will potentially speed up the development of enzyme technologies. Accurately predicting the optimal pH level for enzyme activity is challenging due to the complex relationship between enzyme structure and function. Gado and colleagues show that a language model can effectively learn the structural and biophysical features to predict the optimal pH for enzyme activity.

Synthetic polymers that contain C–O and C–N linkages in their backbones, such as poly(ethylene terephthalate) (PET), polyamides, and polyurethanes, are potential substrates for enzymatic recycling of waste plastics. Enabled by advances in machine learning tools and high-throughput expression, purification, and assaying pipelines for new enzymes, this talk will cover our efforts in enzyme discovery, engineering, and design for the depolymerization of these three classes of waste plastics. Consideration of process economics, energy use, and greenhouse gas emissions (and other environmental impacts consierations) will also be discussed, with the intention that enzyme engineering efforts can focus on the most process-relevant challenges. Overall, the depolymerization of waste plastics that contain backbone C–O and C–N bonds to recoverable and recyclable building blocks is a potentially promising strategy to contribute towards plastics circularity.

Successes in biocatalytic polyester recycling have raised the possibility of deconstructing alternative polymers enzymatically, with polyamide (PA) being a logical target due to the array of amide-cleaving enzymes present in nature. Here, we screen 40 potential natural and engineered nylon-hydrolyzing enzymes (nylonases), using mass spectrometry to quantify eight compounds resulting from enzymatic nylon-6 (PA6) hydrolysis. Comparative time-course reactions incubated at 40-70 °C showcase enzyme-dependent variations in product distributions and extent of PA6 film depolymerization, with significant nylon deconstruction activity appearing rare. The most active nylonase, a NylC K variant we rationally thermostabilized (an N-terminal nucleophile (Ntn) hydrolase, NylC K -TS, T m = 87.4 °C, 16.4 °C higher than the wild-type), hydrolyzes 0.67 wt% of a PA6 film. Reactions fail to restart after fresh enzyme addition, indicating that substrate-based limitations, such as restricted enzyme access to hydrolysable bonds, prohibit more extensive deconstruction. Overall, this study expands our understanding of nylonase activity distribution, indicates that Ntn hydrolases may have the greatest potential for further development, and identifies key targets for progressing PA6 enzymatic depolymerization, including improving enzyme activity, product selectivity, and enhancing polymer accessibility.

Enzymatic deconstruction of poly(ethylene terephthalate) (PET) is under intense investigation, given the ability of hydrolase enzymes to depolymerize PET to its constituent monomers near the polymer glass transition temperature. To date, reported PET hydrolases have been sourced from a relatively narrow sequence space. Here, we identify additional PET-active biocatalysts from natural diversity by using bioinformatics and machine learning to mine 74 putative thermotolerant PET hydrolases. We successfully express, purify, and assay 51 enzymes from seven distinct phylogenetic groups; observing PET hydrolysis activity on amorphous PET film from 37 enzymes in reactions spanning pH from 4.5–9.0 and temperatures from 30–70 °C. We conduct PET hydrolysis time-course reactions with the best-performing enzymes, where we observe differences in substrate selectivity as function of PET morphology. We employed X-ray crystallography and AlphaFold to examine the enzyme architectures of all 74 candidates, revealing protein folds and accessory domains not previously associated with PET deconstruction. Overall, this study expands the number and diversity of thermotolerant scaffolds for enzymatic PET deconstruction.

Cellobiohydrolases (CBHs) in the glycoside hydrolase family 7 (GH7) (EC3.2.1.176) are the major cellulose degrading enzymes both in industrial settings and in the context of carbon cycling in nature. Small carbohydrate conjugates such as p‐nitrophenyl‐β‐d‐cellobioside (pNPC), p‐nitrophenyl‐β‐d‐lactoside (pNPL) and methylumbelliferyl‐β‐d‐cellobioside have commonly been used in colorimetric and fluorometric assays for analysing activity of these enzymes. Despite the similar nature of these compounds the kinetics of their enzymatic hydrolysis vary greatly between the different compounds as well as among different enzymes within the GH7 family. Through enzyme kinetics, crystallographic structure determination, molecular dynamics simulations, and fluorometric binding studies using the closely related compound o‐nitrophenyl‐β‐d‐cellobioside (oNPC), in this work we examine the different hydrolysis characteristics of these compounds on two model enzymes of this class, TrCel7A from Trichoderma reesei and PcCel7D from Phanerochaete chrysosporium. Protein crystal structures of the E212Q mutant of TrCel7A with pNPC and pNPL, and the wildtype TrCel7A with oNPC, reveal that non‐productive binding at the product site is the dominating binding mode for these compounds. Enzyme kinetics results suggest the strength of non‐productive binding is a key determinant for the activity characteristics on these substrates, with PcCel7D consistently showing higher turnover rates (kcat) than TrCel7A, but higher Michaelis–Menten (KM) constants as well. Furthermore, oNPC turned out to be useful as an active‐site probe for fluorometric determination of the dissociation constant for cellobiose on TrCel7A but could not be utilized for the same purpose on PcCel7D, likely due to strong binding to an unknown site outside the active site.

Family 7 glycoside hydrolases (GH7) are among the principal enzymes for cellulose degradation in nature and industrially. These enzymes are often bimodular, including a catalytic domain and carbohydrate-binding module (CBM) attached via a flexible linker, and exhibit an active site that binds cello-oligomers of up to ten glucosyl moieties. GH7 cellulases consist of two major subtypes: cellobiohydrolases (CBH) and endoglucanases (EG). Despite the critical importance of GH7 enzymes, there remain gaps in our understanding of how GH7 sequence and structure relate to function. Here, we employed machine learning to gain data-driven insights into relationships between sequence, structure, and function across the GH7 family. Machine-learning models, trained only on the number of residues in the active-site loops as features, were able to discriminate GH7 CBHs and EGs with up to 99% accuracy, demonstrating that the lengths of loops A4, B2, B3, and B4 strongly correlate with functional subtype across the GH7 family. Classification rules were derived such that specific residues at 42 different sequence positions each predicted the functional subtype with accuracies surpassing 87%. A random forest model trained on residues at 19 positions in the catalytic domain predicted the presence of a CBM with 89.5% accuracy. Our machine learning results recapitulate, as top-performing features, a substantial number of the sequence positions determined by previous experimental studies to play vital roles in GH7 activity. We surmise that the yet-to-be-explored sequence positions among the top-performing features also contribute to GH7 functional variation and may be exploited to understand and manipulate function.

Significance Deconstruction of recalcitrant polymers, such as cellulose or chitin, is accomplished in nature by synergistic enzyme cocktails that evolved over millions of years. In these systems, soluble dimeric or oligomeric intermediates are typically released via interfacial biocatalysis, and additional enzymes often process the soluble intermediates into monomers for microbial uptake. The recent discovery of a two-enzyme system for polyethylene terephthalate (PET) deconstruction, which employs one enzyme to convert the polymer into soluble intermediates and another enzyme to produce the constituent PET monomers (MHETase), suggests that nature may be evolving similar deconstruction strategies for synthetic plastics. This study on the characterization of the MHETase enzyme and synergy of the two-enzyme PET depolymerization system may inform enzyme cocktail-based strategies for plastics upcycling.

Accurate prediction of the optimal catalytic temperature (Topt) of enzymes is vital in biotechnology, as enzymes with high Topt values are desired for enhanced reaction rates. Recently, a machine-learning method (TOME) for predicting Topt was developed. TOME was trained on a normally-distributed dataset with a median Topt of 37°C and less than five percent of Topt values above 85°C, limiting the method’s predictive capabilities for thermostable enzymes. Due to the distribution of the training data, the mean squared error on Topt values greater than 85°C is nearly an order of magnitude higher than the error on values between 30 and 50°C. In this study, we apply ensemble learning and resampling strategies that tackle the data imbalance to significantly decrease the error on high Topt values (>85°C) by 60% and increase the overall R2 value from 0.527 to 0.632. The revised method, TOMER, and the resampling strategies applied in this work are freely available to other researchers as a Python package on GitHub.

Microbial conversion of aromatic compounds is an emerging and promising strategy for valorization of the plant biopolymer lignin. A critical and often rate-limiting reaction in aromatic catabolism is O-aryl-demethylation of the abundant aromatic methoxy groups in lignin to form diols, which enables subsequent oxidative aromatic ring-opening. Recently, a cytochrome P450 system, GcoAB, was discovered to demethylate guaiacol (2-methoxyphenol), which can be produced from coniferyl alcohol-derived lignin, to form catechol. However, native GcoAB has minimal ability to demethylate syringol (2,6-dimethoxyphenol), the analogous compound that can be produced from sinapyl alcohol-derived lignin. Despite the abundance of sinapyl alcohol-based lignin in plants, no pathway for syringol catabolism has been reported to date. Here we used structure-guided protein engineering to enable microbial syringol utilization with GcoAB. Specifically, a phenylalanine residue (GcoA-F169) interferes with the binding of syringol in the active site, and on mutation to smaller amino acids, efficient syringol O-demethylation is achieved. Crystallography indicates that syringol adopts a productive binding pose in the variant, which molecular dynamics simulations trace to the elimination of steric clash between the highly flexible side chain of GcoA-F169 and the additional methoxy group of syringol. Finally, we demonstrate in vivo syringol turnover in Pseudomonas putida KT2440 with the GcoA-F169A variant. Taken together, our findings highlight the significant potential and plasticity of cytochrome P450 aromatic O-demethylases in the biological conversion of lignin-derived aromatic compounds.