mLife专题分享 | 微生物工程研究

Biosynthesis—a process utilizing biological systems to synthesize chemical compounds—has emerged as a revolutionary solution to 21st-century challenges due to its environmental sustainability, scalability, and high stereoselectivity and regioselectivity. Recent advancements in artificial intelligence (AI) are accelerating biosynthesis by enabling intelligent design, construction, and optimization of enzymatic reactions and biological systems. We first introduce the molecular retrosynthesis route planning in biochemical pathway design, including single-step retrosynthesis algorithms and AI-based chemical retrosynthesis route design tools. We highlight the advantages and challenges of large language models in addressing the sparsity of chemical data. Furthermore, we review enzyme discovery methods based on sequence and structure alignment techniques. Breakthroughs in AI-based structural prediction methods are expected to significantly improve the accuracy of enzyme discovery. We also summarize methods for de novo enzyme generation for nonnatural or orphan reactions, focusing on AI-based enzyme functional annotation and enzyme discovery techniques based on reaction or small molecule similarity. Turning to enzyme engineering, we discuss strategies to improve enzyme thermostability, solubility, and activity, as well as the applications of AI in these fields. The shift from traditional experiment-driven models to data-driven and computationally driven intelligent models is already underway. Finally, we present potential challenges and provide a perspective on future research directions. We envision expanded applications of biocatalysis in drug development, green chemistry, and complex molecule synthesis.

Advances in deep learning have significantly aided protein engineering in addressing challenges in industrial production, healthcare, and environmental sustainability. This review frames frequently researched problems in protein understanding and engineering from the perspective of deep learning. It provides a thorough discussion of representation methods for protein sequences and structures, along with general encoding pipelines that support both pre-training and supervised learning tasks. We summarize state-of-the-art protein language models, geometric deep learning techniques, and the combination of distinct approaches to learning from multi-modal biological data. Additionally, we outline common downstream tasks and relevant benchmark datasets for training and evaluating deep learning models, focusing on satisfying the particular needs of protein engineering applications, such as identifying mutation sites and predicting properties for candidates'' virtual screening. This review offers biologists the latest tools for assisting their engineering projects while providing a clear and comprehensive guide for computer scientists to develop more powerful solutions by standardizing problem formulation and consolidating data resources. Future research can foresee a deeper integration of the communities of biology and computer science, unleashing the full potential of deep learning in protein engineering and driving new scientific breakthroughs.

The yeast Saccharomyces cerevisiae is a well-studied unicellular eukaryote with a significant role in the biomanufacturing of natural products, biofuels, and bulk and value-added chemicals, as well as the principal model eukaryotic organism utilized for fundamental research. Robust tools for building and optimizing yeast chassis cells were made possible by the quick development of synthetic biology, especially in engineering evolution. In this review, we focused on methods and tools from synthetic biology that are used to design and engineer S. cerevisiae''s evolution. A detailed discussion was held regarding transcriptional regulation, template-dependent and template-free approaches. Furthermore, the applications of evolved S. cerevisiae were comprehensively summarized. These included improving environmental stress tolerance and raising cell metabolic performance in the production of biofuels and bulk and value-added chemicals. Finally, the future considerations were briefly discussed.

Optimizing enzyme thermostability is essential for advancements in protein science and industrial applications. Currently, (semi-)rational design and random mutagenesis methods can accurately identify single-point mutations that enhance enzyme thermostability. However, complex epistatic interactions often arise when multiple mutation sites are combined, leading to the complete inactivation of combinatorial mutants. As a result, constructing an optimized enzyme often requires repeated rounds of design to incrementally incorporate single mutation sites, which is highly time-consuming. In this study, we developed an AI-aided strategy for enzyme thermostability engineering that efficiently facilitates the recombination of beneficial single-point mutations. We utilized thermostability data from creatinase, including 18 single-point mutants, 22 double-point mutants, 21 triple-point mutants, and 12 quadruple-point mutants. Using these data as inputs, we used a temperature-guided protein language model, Pro-PRIME, to learn epistatic features and design combinatorial mutants. After two rounds of design, we obtained 50 combinatorial mutants with superior thermostability, achieving a success rate of 100%. The best mutant, 13M4, contained 13 mutation sites and maintained nearly full catalytic activity compared to the wild-type. It showed a 10.19°C increase in the melting temperature and an ~655-fold increase in the half-life at 58°C. Additionally, the model successfully captured epistasis in high-order combinatorial mutants, including sign epistasis (K351E) and synergistic epistasis (D17V/I149V). We elucidated the mechanism of long-range epistasis in detail using a dynamics cross-correlation matrix method. Our work provides an efficient framework for designing enzyme thermostability and studying high-order epistatic effects in protein-directed evolution.

In silico computational methods have been widely utilized to study enzyme catalytic mechanisms and design enzyme performance, including molecular docking, molecular dynamics, quantum mechanics, and multiscale QM/MM approaches. However, the manual operation associated with these methods poses challenges for simulating enzymes and enzyme variants in a high-throughput manner. We developed the NAC4ED, a high-throughput enzyme mutagenesis computational platform based on the “near-attack conformation” design strategy for enzyme catalysis substrates. This platform circumvents the complex calculations involved in transition-state searching by representing enzyme catalytic mechanisms with parameters derived from near-attack conformations. NAC4ED enables the automated, high-throughput, and systematic computation of enzyme mutants, including protein model construction, complex structure acquisition, molecular dynamics simulation, and analysis of active conformation populations. Validation of the accuracy of NAC4ED demonstrated a prediction accuracy of 92.5% for 40 mutations, showing strong consistency between the computational predictions and experimental results. The time required for automated determination of a single enzyme mutant using NAC4ED is 1/764th of that needed for experimental methods. This has significantly enhanced the efficiency of predicting enzyme mutations, leading to revolutionary breakthroughs in improving the performance of high-throughput screening of enzyme variants. NAC4ED facilitates the efficient generation of a large amount of annotated data, providing high-quality data for statistical modeling and machine learning. NAC4ED is currently available at http://lujialab.org.cn/software/.

Inflammatory bowel disease (IBD) is a chronic condition characterized by intestinal inflammation and gut dysbiosis, with limited treatment options primarily focused on immune-modulating therapies. Among potential therapeutic agents, butyrate has emerged as a promising candidate due to its anti-inflammatory and gut-restorative properties. However, direct administration of butyrate poses significant challenges, including its rapid absorption, uneven distribution within the intestinal tract, and an unpleasant odor that reduces patient compliance. To address these issues, we evaluated the therapeutic potential of Bacillus subtilis BM107, a strain selected for its superior butyrate-producing capabilities and established bacterial safety. BM107 efficiently hydrolyzed tributyrin (TB), a butyrate prodrug, producing substantial butyrate levels in TB-supplemented media. In a dextran sodium sulfate-induced colitis mouse model, co-administration of BM107 and the TB diet significantly improved inflammatory indices, such as reduced disease activity index scores, increased colon length, and restored body weight. Additionally, this combination treatment markedly improved gut microbiome composition, restoring microbial diversity and balance. Furthermore, butyrate levels in the cecum contents of the TB + BM107 group were restored to levels comparable to those in healthy controls, demonstrating the ability of this approach to promote gut homeostasis and intestinal recovery. These findings highlight the therapeutic potential of BM107 combined with a TB diet as a safe, effective, and innovative strategy for addressing gut dysbiosis and inflammation in IBD, paving the way for the development of microbiome-based bacterial therapeutics to improve patient outcomes.

Halophilic proteins possess unique structural properties and show high stability under extreme conditions. This distinct characteristic makes them invaluable for application in various aspects such as bioenergy, pharmaceuticals, environmental clean-up, and energy production. Generally, halophilic proteins are discovered and characterized through labor-intensive and time-consuming wet lab experiments. In this study, we introduce the Halophilic Protein Classifier (HPClas), a machine learning-based classifier developed using the catBoost ensemble learning technique to identify halophilic proteins. Extensive in silico calculations were conducted on a large public dataset of 12,574 samples and HPClas achieved an area under the receiver operating characteristic curve (AUROC) of 0.844 on an independent test set of 200 samples. The source code and curated dataset of HPClas are publicly available at https://github.com/Showmake2/HPClas. In conclusion, HPClas can be explored as a promising tool to aid in the identification of halophilic proteins and accelerate their application in different fields.

Kluyveromyces marxianus is a food-safe yeast with great potential for producing heterologous proteins. Improving the yield in K. marxianus remains a challenge and incorporating large-scale functional modules poses a technical obstacle in engineering. To address these issues, linear and circular yeast artificial chromosomes of K. marxianus (KmYACs) were constructed and loaded with disulfide bond formation modules from Pichia pastoris or K. marxianus. These modules contained up to seven genes with a maximum size of 15 kb. KmYACs carried telomeres either from K. marxianus or Tetrahymena. KmYACs were transferred successfully into K. marxianus and stably propagated without affecting the normal growth of the host, regardless of the type of telomeres and configurations of KmYACs. KmYACs increased the overall expression levels of disulfide bond formation genes and significantly enhanced the yield of various heterologous proteins. In high-density fermentation, the use of KmYACs resulted in a glucoamylase yield of 16.8 g/l, the highest reported level to date in K. marxianus. Transcriptomic and metabolomic analysis of cells containing KmYACs suggested increased flavin adenine dinucleotide biosynthesis, enhanced flux entering the tricarboxylic acid cycle, and a preferred demand for lysine and arginine as features of cells overexpressing heterologous proteins. Consistently, supplementing lysine or arginine further improved the yield. Therefore, KmYAC provides a powerful platform for manipulating large modules with enormous potential for industrial applications and fundamental research. Transferring the disulfide bond formation module via YACs proves to be an efficient strategy for improving the yield of heterologous proteins, and this strategy may be applied to optimize other microbial cell factories.

Hyperthermus butylicus is a hyperthermophilic crenarchaeon that produces 1-butanol as an end product. A thermostable alcohol dehydrogenase (ADH) must be present in H. butylicus to act as the key enzyme responsible for this production; however, the gene that encodes the ADH has not yet been identified. A novel ADH, HbADH2, was purified from a cell-free extract of H. butylicus, and its characteristics were determined. The gene that encodes HbADH2 was demonstrated to be HBUT_RS04850 and annotated as a hypothetical protein in H. butylicus. HbADH2 was found to be a primary–secondary ADH capable of using a wide range of substrates, including butyraldehyde and butanol. Butyraldehyde had the highest specificity constant, calculated as kcat/Km, with kcat and apparent Km values of 8.00 ± 0.22 s−1 and 0.59 ± 0.07 mM, respectively. The apparent Km values for other substrates, including ethanol, 1-propanol, 2-propanol, butanol, acetaldehyde, propanal, and acetone, were 4.36 ± 0.42, 4.69 ± 0.41, 3.74 ± 0.46, 2.44 ± 0.30, 1.27 ± 0.18, 1.55 ± 0.20, and 0.68 ± 0.04 mM, respectively. The optimal pH values for catalyzing aldehyde reduction and alcohol oxidation were 6.0 and 9.0, respectively, while the optimal temperature was higher than 90°C due to the increase in enzymatic activity from 60°C to 90°C. Based on its substrate specificity, enzyme kinetics, and thermostability, HbADH2 may be the ADH that catalyzes the production of 1-butanol in H. butylicus. The putative conserved motif sites for NAD(P)+ and iron binding were identified by aligning HbADH2 with previously characterized Fe-containing ADHs.

Saccharolobus islandicus REY15A represents one of the very few archaeal models with versatile genetic tools, which include efficient genome editing, gene silencing, and robust protein expression systems. However, plasmid vectors constructed for this crenarchaeon thus far are based solely on the pRN2 cryptic plasmid. Although this plasmid coexists with pRN1 in its original host, early attempts to test pRN1-based vectors consistently failed to yield any stable host–vector system for Sa. islandicus. We hypothesized that this failure could be due to the occurrence of CRISPR immunity against pRN1 in this archaeon. We identified a putative target sequence in orf904 encoding a putative replicase on pRN1 (target N1). Mutated targets (N1a, N1b, and N1c) were then designed and tested for their capability to escape the host CRISPR immunity by using a plasmid interference assay. The results revealed that the original target triggered CRISPR immunity in this archaeon, whereas all three mutated targets did not, indicating that all the designed target mutations evaded host immunity. These mutated targets were then incorporated into orf904 individually, yielding corresponding mutated pRN1 backbones with which shuttle plasmids were constructed (pN1aSD, pN1bSD, and pN1cSD). Sa. islandicus transformation revealed that pN1aSD and pN1bSD were functional shuttle vectors, but pN1cSD lost the capability for replication. These results indicate that the missense mutations in the conserved helicase domain in pN1c inactivated the replicase. We further showed that pRN1-based and pRN2-based vectors were stably maintained in the archaeal cells either alone or in combination, and this yielded a dual plasmid system for genetic study with this important archaeal model.

We have developed the GReedy Accumulated strategy for Protein Engineering (GRAPE) to improve enzyme stability across various applications, combining advanced computational methods with a unique clustering and greedy accumulation approach to efficiently explore epistatic effects with minimal experimental effort. To make this strategy accessible to nonexperts, we introduced GRAPE-WEB, an automated, user-friendly web server that allows the design, inspection, and combination of stabilizing mutations without requiring extensive bioinformatics knowledge. GRAPE-WEB''s robust performance and accessibility provide a comprehensive and adaptable approach to protein thermostability design, suitable for both newcomers and experienced practitioners in the field. The web server is accessible at https://grape.wulab.xyz.