Drug target affinity (DTA) prediction is a critical component of virtual screening and plays a pivotal role in accelerating drug discovery. However, existing deep learning-based DTA methods often rely on unimodal representations (such as textual or graph-based embeddings), which are insufficient to capture the complex interactions between drugs and targets. To address this limitation, a cross-modal fusion framework, CMF-DTA, is proposed that independently integrates drug text expressions with molecular graphs and target protein amino acid sequences with protein structure graphs. This strategy enhances the diversity and expressiveness of feature representations. To overcome the limitations of the conventional graph isomorphism network (GIN) in neighborhood aggregation, which restricts the representation of structural information, an extended second-order neighborhood self-attention aggregation mechanism is designed. This mechanism adaptively captures both first- and second-order neighborhood features, improving structural awareness. Furthermore, a dual-guided attention mechanism facilitates efficient cross-modal interaction and fusion of text and graph modality features, exploiting complementary information across modalities. Experimental results demonstrate that CMF-DTA achieves state-of-the-art performance on benchmark datasets, exhibiting strong generalization and resilience.
Clostridium carboxidivorans P7 can utilize syngas as a carbon and energy source to synthesize solvents (ethanol and butanol), but the function of genes that play a crucial role in alcohol synthesis has not yet been elucidated. Bioinformatics analysis showed that the genome of C. carboxidivorans P7 contained two genes, adhE1 and adhE2 (Ccar_07995 and Ccar_08000), encoding bifunctional alcohol and aldehyde dehydrogenase, whose amino acid sequences were 63% and 79% identical to those of the bifunctional alcohol and aldehyde dehydrogenase from the model Escherichia coli, respectively. This study characterizes AdhE1 and AdhE2 of C. carboxidivorans P7 through in vitro and in vivo experiments. Enzymatic characterization analysis demonstrated that both enzymes are capable of catalyzing the synthesis of ethanol from acetyl-CoA and butanol from butyryl-CoA. The enzyme activities of AdhE1 and AdhE2 in catalyzing the conversion of acetyl-CoA to ethanol are 0.40 U/mg and 1.20 U/mg, respectively. Meanwhile, their enzyme activities in catalyzing the conversion of butyryl-CoA to butanol are 0.26 U/mg and 1.20 U/mg, respectively, which indicates that the overall catalytic activity of AdhE2 is significantly higher than that of AdhE1. In addition, by overexpressing adhE1 and adhE2, driven by the constitutive promoter Ppta, in the model syngas-utilizing species Clostridium ljungdahlii in vivo, we found that the ability of AdhE1 and AdhE2 overexpressing strains to synthesize ethanol from syngas was significantly enhanced, with 50% and 75% increases in ethanol titer, respectively. The above study indicates that adhE1 and adhE2 are key alcohol synthesis genes in C. carboxidivorans P7. This research provides important clues to elucidate the key mechanisms of alcohol synthesis from syngas in C. carboxidivorans P7 at the molecular level.
The clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nuclease 9 (Cas9) system represents a powerful and precise gene-editing tool. However, its widespread clinical application is hampered by the lack of safe, tissue-specific, and efficient delivery systems. A critical challenge for CRISPR/Cas9-based therapeutics lies in achieving safe and efficient in vivo delivery to target tissues. Exosomes, nanometer-sized extracellular vesicles, play a pivotal role in intercellular communication and facilitate the transfer of biomolecules, including proteins and nucleic acids. In this study, Cas9/single guide RNA (sgRNA)-ribonucleoprotein (RNP) complexes were successfully encapsulated in exosomes by heterodimerization of Cas9 fused to the N-terminus of the exosomal membrane protein Lamp2b coupled to the hepatocyte-targeting polypeptide SP94. These engineered exosomes, termed Cas9/sgRNA@Exo, exhibited a cup-shaped morphology with an average diameter of approximately 125 nm. Functional assays demonstrated that Cas9/sgRNA@Exo effectively delivered functional Cas9 and sgRNA to Huh7-eGFP cells, enabling targeted cleavage of the enhanced green fluorescent protein (eGFP) gene. Importantly, these engineered exosomes demonstrated excellent biocompatibility and safety. In conclusion, this study has established a robust and safe exosome-based platform for liver-specific gene therapy, offering a promising strategy for targeted CRISPR/Cas9 delivery in clinical applications.
Objective: The study was designed to develop a high-yield production system for low molecular weight poly-γ-glutamic acid (LMW-γ-PGA) by screening poly-γ-glutamic acid synthase (PgsBCA) and depolymerase (PgdS) from diverse microbial sources, followed by their co-expression in Corynebacterium glutamicum. Additionally, a promoter engineering strategy was proposed to eliminate chemical inducer dependence and enhance the coordinated expression of PgsBCA and PgdS to obtain high-performance LMW-γ-PGA-producing strains. Methods: Three PgsBCA homologs were heterologously expressed in C. glutamicum, and the optimal variant was identified by fermentation performance evaluation. Subsequently, three candidate PgdS were co-expressed with the selected PgsBCA and strain performance was evaluated to determine the superior enzyme combination. Promoter engineering was implemented by screening native promoters and integrating constitutive promoters of varying strength to establish an inducer-free system for balanced PgsBCA-PgdS co-expression and enhanced LMW-γ-PGA biosynthesis. Results: Co-expression of Bacillus licheniformis PgsBCA and Bacillus spizizenii PgdS in C. glutamicum yielded 14.5 g/L γ-PGA with a molecular weight ≤200 kDa. Promoter engineering identified Pdep-A16 as the optimal replacement for the inducible Ptac, enabling inducer-free LMW-γ-PGA production. A 15 L fed-batch fermentation demonstrated scalable production, yielding 37.42 g/L LMW-γ-PGA at a productivity of 0.78 g/(L·h). Conclusion: We established an optimized PgsBCA-PgdS co-expression system in C. glutamicum and developed an inducer-free process through promoter engineering, enabling high-yield LMW-γ-PGA production in industrial-scale fermentation.
Recombinant protein expression is a pivotal aspect of bioengineering research and application, and the achievement of high-level and high-activity protein expression has significant theoretical and practical value. For a long time, the optimization of recombinant protein expression levels has focused primarily on examining and refining external factors such as the composition of the recombinant expression system, the culture conditions of the recombinant engineering bacteria, and the induction conditions. In contrast, research on internal factors such as protein coding sequences and regulatory sequences has been relatively limited. With the rapid development and deep integration of bioinformatics, computational biology, and artificial intelligence technologies, there have been increasing efforts to utilize machine learning methods to construct models for predicting protein expression levels or generating protein sequences. These models are then used for intelligent optimization or de novo design of protein sequences. This strategy not only enhances the precision and efficiency of the optimization of recombinant protein expression sequences, but also provides new research perspectives for a deeper understanding of the regulatory mechanisms of protein expression levels, thereby attracting considerable attention from researchers. This paper summarizes recent advances in the application of machine learning to the optimization and design of recombinant protein expression sequences. It focuses on analyzing and discussing the construction methods of representative machine learning models and sequence optimization strategies, as well as addressing key challenges such as overcoming the limitations of training data and improving model interpretability.
Targeted therapies have been a revolutionary breakthrough in cancer treatment. A key challenge in this area is the efficient delivery of therapeutic agents to cancerous tissues while ensuring enhanced interaction with specific cellular targets. Although various nanoparticle carriers have been developed to facilitate targeted drug delivery, the lysosomal trap remains a significant obstacle, preventing many drugs from effectively reaching their intracellular targets. Membrane fusion, a fundamental biological process, plays a crucial role in the renewal of cellular membrane lipids and the delivery of vesicular contents. By mimicking this natural fusion mechanism, targeted drug delivery to cancer tissues and effective interaction with intracellular targets can be achieved. Given recent advances in the field and the lack of comprehensive reviews, this article reviews the principles and methods of membrane fusion and its role in enabling efficient interactions between anticancer drugs and their targets. Additionally, it reviews the latest applications, current challenges, and potential developments of membrane fusion technologies in cancer treatment.
Based on the specific affinity between sugars and proteins, sugars have been widely used as key affinity ligands in the construction of glycosyl-modified functionalized interfaces. Compared with affinity ligands such as proteins and nucleic acids, sugars have a wide range of sources, high biocompatibility, low cost and strong environmental stability, and have attracted much attention in the fields of biomedicine, food science and environmental science in recent years. Biomimetic construction of functionalized polyglycosyl-modified affinity interface can significantly enhance the binding strength between the interface and the receptor, overcome the off-target and non-specific adsorption problems caused by weak monovalent glycosyl binding force, and show excellent physicochemical properties in improving the detection sensitivity and separation efficiency and realizing accurate drug delivery. According to the different spatial dimensions of glycosylation-modified carriers, the unique biochemical properties of zero-dimensional (gold nanoparticles, silver nanoparticles, and iron oxide nanoparticles), one-dimensional (carbon nanotubes), two-dimensional (graphene oxide) and three-dimensional (antimicrobial peptides, liposomes, and dendritic macromolecules) carriers were introduced. The latest research progress of different glycosyl-modified functionalized interfaces in the fields of drug targeting, pathogen detection, disease diagnosis, etc. was reviewed, and the shortcomings and development prospects of these interfaces were examined. It is expected to provide data support for the research and development of new glycosyl-modified functionalized interfaces, and provide an important basis for a comprehensive understanding of the research trends of bionic functionalized affinity interfaces in food safety detection, environmental monitoring, biomedicine and other fields.
Bone defect refers to the complex pathological changes of partial or total loss of bone tissue caused by trauma, bone tumor resection, infection or other causes. With the development of materials science and biotechnology, biomaterials are widely used in bone tissue engineering to repair bone defects. Porous biological scaffolds are an important branch of biomaterials, and their application in bone defect repair has become one of the important research directions of regenerative medicine. Osteogenesis-angiogenesis coupling refers to the interaction and synergistic process between bone formation and angiogenesis, which is essential for normal bone development, regeneration and repair. The pores of the porous biological scaffold, designed based on this theory, allow vascular endothelial cells to enter and form new blood vessels. At the same time, the material can release growth factors and enhance the synergistic effect between vascular endothelial cells and osteoblasts. This coupling mechanism ensures that new bone tissue receives sufficient nutrients and oxygen in the scaffold, accelerating bone regeneration and improving the repair effect. However, at present, the classification of porous biological scaffolds is vague, and the screening of materials is time-consuming and laborious. It is difficult for clinicians to accurately select appropriate scaffolds for patient treatment. Therefore, this paper summarizes the application of porous biological scaffolds based on the osteogenesis-angiogenesis coupling mechanism in bone repair, aiming to help researchers and clinicians better understand the design principles and applicable scenarios of different types of scaffolds, and then to achieve more efficient and stable bone defect repair in clinical practice.
Biocontrol bacteria refer to beneficial microorganisms that can be used to prevent and control plant diseases. Their application in agricultural control is becoming increasingly widespread, providing new ideas for cost-effective and efficient green prevention and control of plant diseases. However, there is still a gap in the efficacy of biocontrol agents compared with traditional chemical pesticides, and large-scale commercial production has rarely been achieved. In addition, the lack of a systematic and comprehensive review does not allow researchers to intuitively understand the latest research hotspots. This article reviews the research and application status of biocontrol fermentation based on major databases in the past 10 years, classifies the types of fermentation processes and related important parameters, lists common optimization methods, and discusses the strategies and practical applications for constructing high-yielding biocontrol engineering bacteria from a genetic perspective. Meanwhile, this article examines the current problems with biopesticides and suggests possible solutions. Finally, the application prospects and challenges faced in the future development of biocontrol bacteria are discussed from two aspects: macro fermentation process optimization and molecular regulation construction of engineering bacteria.
As a key enzyme in the coenzyme A biosynthetic pathway, pantothenic acid synthase (PS) plays an indispensable role in various life processes. This enzyme is involved in the regulation of basic physiological activities such as protein ubiquitination, lipid metabolism, and cellular energy production. Although PS is widely distributed in the biological world, it is interesting to note that PS derived from prokaryotes and eukaryotes exhibit significant differences in sequence, with homology typically not exceeding 20%. More importantly, normal bacterial growth is highly dependent on PS-mediated pantothenic acid synthesis, making PS an attractive target for the development of new antibacterial drugs.This review aims to provide a comprehensive overview of the latest developments in PS-related research. We focus on the following aspects. First, we explore the structural properties of PS and its correlation with function. Second, we conduct a thorough analysis of the catalytic mechanism of PS and summarize the research results of directed evolution of PS in the field of protein engineering. Finally, an overview of the current status of inhibitor development for PS is reviewed. On this basis, we also discuss the future research directions and potential applications of PS in order to provide valuable reference and inspiration for researchers in this field.
Equol is a product of daidzein that is metabolized by intestinal microorganisms through a series of enzymes. It has many biological activities such as estrogen-like, antioxidant and anti-cancer activities and has important research value in the prevention and treatment of a variety of diseases. Biosynthesis is the most promising method for the preparation of equol. This paper reviews the research status of four key enzymes in the equol biosynthetic pathway, namely daidzein reductase DZNR, dihydrodaidzein reductase DHDR, tetrahydrodaidzein reductase THDR, and dihydrodaidzein racemase DDRC. By exploring the mechanism of action of these key enzymes, we can provide important theoretical support and practical guidance for the biosynthesis and yield improvement of equol, thus promoting the wide application of equol in various related fields.
As a frontier discipline of the 21st century, synthetic biology is reshaping the research paradigm in the life sciences with its revolutionary technological tools. As a model plant and an important industrial raw material, tobacco has become a crucial platform for synthetic biology research due to its complete secondary metabolic network and unique biosynthetic capabilities. In-depth exploration of tobacco's biosynthetic potential has significant theoretical and practical value for the development of novel drugs, crop trait improvement, and sustainable production. This paper systematically reviews recent research progress in tobacco synthetic biology, focusing on the construction and application of tobacco bioreactors, molecular mechanisms of nicotine metabolism regulation, research on key enzyme systems for terpene biosynthesis, and the discovery of active metabolites from endophytic fungi and their prospects for application in the synthesis of medicinal compounds such as paclitaxel. In reviewing these studies, it is clear that synthetic biology is driving the transformation of tobacco from a traditional crop to a multifunctional biofactory, a transformation that will make significant contributions to human health and sustainable development.