ACS Synthetic Biology最新文献

Protein liquid-liquid phase separation underlies the formation of membraneless organelles in cells and plays a key role in the assembly process of natural materials such as the assembly of tropoelastin into elastic fibers. Here, we engineered a series of charged elastin-like polypeptides (ELPs) that form complex coacervates, providing a rapid method of concentrating proteins into a fluid state. Compared with coacervates formed via simple coacervation, complex coacervates exhibited greater fluidity, likely due to differences between electrostatic interactions and hydrophobic forces. We designed these ELPs to further contain cross-linking domains compatible with tyrosinase or transglutaminase and found that cross-linking was enhanced when proteins were in a condensed state compared to free in solution. Cross-linking the ELP complex coacervates led to the formation of gels with distinct properties dependent on the nature of the cross-linking. This work expands the design space of protein hydrogels, offering a novel strategy for forming cross-linked networks from complex coacervates and providing opportunity for future use in tissue engineering and biocompatible biomaterials applications.

Combination therapy is widely used in clinical practice, rendering accurate prediction of drug-drug interactions (DDIs) essential for treatment safety and efficacy. This paper proposes PTET-DDI, a novel dual-channel framework that synergizes chemical semantics with geometric structural insights for DDI prediction. Unlike existing approaches that rely solely on 1D sequences or 2D graphs, PTET-DDI integrates a pretrained molecular language model (ChemBERTa) to capture rich context-aware semantic representations, while simultaneously employing an improved fully equivariant graph Transformer to explicitly encode 3D molecular conformations and geometric symmetries. By fusing these complementary modalities, the model not only achieves a comprehensive understanding of molecular properties but also provides interpretability by identifying key structures driving the interactions. Extensive evaluations via 5-fold cross-validation across three benchmark data sets demonstrate that PTET-DDI significantly outperforms existing deep learning-based methods and shows strong generalization capability.

Recombinant protein expression in mycobacteria faces two major challenges: limited regulatory tools for inducible expression and inefficient secretion of heterologous products. In this study, we developed plasmid-based systems that enable translationally gated secretion in Mycobacterium smegmatis, coupling riboswitch-mediated translational control with efficient extracellular export. The platform integrates the M. tuberculosis antigen 85A promoter and signal peptide for constitutive secretion combined with synthetic riboswitches for inducible translational regulation. We tested two theophylline-responsive riboswitches (riboE and riboE+) and a temperature-sensitive variant (riboU9) by using mCherry as a reporter. Fluorescence assays, RT-PCR, and Western blotting confirmed efficient secretion and strict translational control. The theophylline-inducible systems exhibited a dose-dependent response with maximal expression at 2 mM inducer, while the riboU9 construct showed a clean ON/OFF phenotype triggered by temperature shift. In all cases, transcripts were detected irrespective of induction, confirming regulation at the translational rather than transcriptional level. Secretion was highly efficient, with 10-20 fold higher protein levels in extracellular versus intracellular fractions. Induction during early- and mid-log phases yielded maximal protein, whereas late-log induction reduced output by ∼50%. Together, these results define translationally gated secretion as a new control layer in mycobacterial protein production. This modular platform expands the genetic toolkit available for Mycobacterium research, providing new opportunities for the study of antigens and virulence factors from slow-growing pathogens and offering potential applications in structural biology, vaccine development, and drug target validation.

Cancer is known to be a disease of altered cellular signaling; however, the relationship between mutation-specific changes to signal transduction and the phenotypic consequences produced remains poorly understood. Here, we investigate two common breast cancer driver mutations, the PIK3CA^H1047R mutation and the ErbB2 amplification, both of which activate the PI3K-Akt pathway but paradoxically drive distinct cellular outcomes. Indeed, in nontransformed mammary epithelial cells, PI3K^H1047R expression induced features of epithelial-mesenchymal transition (EMT), while ErbB2^amp cells exhibited a hyperproliferative phenotype. Characterization of PI3K axis signaling revealed that ErbB2^amp cells display prolonged, stimulus-dependent PI3K activation, whereas PI3K^H1047R cells show constitutive, ligand-independent signaling. To test whether these distinct dynamics contribute to the phenotypic responses, we employed an iLID-based optogenetic system that enables precise, tunable control of endogenous PI3K activity. Using this tool to mimic the mutation-specific dynamics in MCF10A mammary epithelial cells, we found that PI3K signaling patterns alone were sufficient to reproduce key features of the PIK3CA H1047R-associated EMT phenotype but not the ErbB2-associated proliferative phenotype. These findings suggest that the temporal encoding of pathway activity, not merely its magnitude, can drive some phenotypic changes in oncogenic progression, explain how distinct mutations within a common signaling pathway can produce divergent cellular phenotypes, and provide a workflow for interrogating the functional consequences of changes in signaling dynamics.

Ethanol is a widely available and inexpensive commodity chemical, and can be utilized by microorganisms with high carbon yield, making it a potential alternative carbon source in fermentation. In this study, fumarate, a versatile platform chemical, was biosynthesized from ethanol via partial tricarboxylic acid (TCA) cycle and glyoxylate shunt in engineered Escherichia coli strain. To address growth deficiency of E. coli on ethanol, adaptive laboratory evolution was performed, resulting in evolved strains with improved growth and enhanced fumarate production. Knockout of citrate lyase and increased adenylate cyclase activity significantly contributed to growth improvement. Furthermore, gluconeogenesis and glyoxylate shunt were further transcriptionally activated in the evolved strain. Finally, the optimized strain Q6926 produced 31.03 mM fumarate from ethanol in fed-batch fermentation. This work represents the first demonstration of growth-coupled fumarate bioproduction from ethanol, demonstrating the feasibility of using ethanol as a sustainable feedstock for the biosynthesis of TCA cycle-derived chemicals.

Internal ribosome entry sites (IRESs) provide compact RNA elements for noncanonical translation and hold promise as building blocks for RNA-based regulation in synthetic biology. However, the cricket paralysis virus (CrPV) IRES shows very low activity in Saccharomyces cerevisiae, limiting its broader utility despite extensive structural and biochemical studies. Here we report a yeast engineering strategy that enhances CrPV IRES-mediated translation by combining host modifications at three mechanistically distinct levels: translation initiation, tRNA modification, and mRNA stability. A reporter-based screen revealed host factors that influence IRES activity and uncovered a trade-off between IRES stimulation and maintenance of cap-dependent translation required for growth. Stepwise integration of nonsense-mediated decay deficiency, a tad3 temperature-sensitive allele, and wild-type eIF4E overexpression yielded a strain with up to an order-of-magnitude increase in reporter output compared with that of the parental strain. These results establish a proof-of-principle framework for host engineering of noncanonical translation.

RNA regulators offer a promising path for building complex, orthogonal circuits due to their low resource demands and design flexibility. In this study, we explore their potential as signaling molecules in communication between synthetic cells. Specifically, we engineer populations of heterogenetic porous polymer cell mimics to produce, emit, and receive two types of small synthetic RNA regulators. These RNAs are required to activate reporter expression at both the levels of transcription and translation. We distribute this AND gate circuit in receiver and two types of sender cell mimics to compare the distributed logic computation to the behavior of the circuit in well-mixed, bulk cell-free expression reactions. Analyzing different densities and spatial arrangements of senders and receivers, we reveal spatiotemporal gradients in RNA signals and identify configurations that increase specific activation. With small regulatory RNAs, the engineering toolbox for communication between synthetic cells expands to include a programmable class of signaling molecules. The rapid turnover of RNA suggests applications in establishing dynamic signaling gradients in communities of synthetic cells.

Mycomaterials, materials made from filamentous fungi, have several advantages over traditional materials such as their genetic programmability and self-healing properties. However, their lack of mechanical strength and cost of production often constrain the applications in which they can be used in. In this work, we take inspiration from natural systems to overcome these challenges by elucidating design principles for mineralization-based enhancement of mechanical strength and synthetic lichen-based low-cost growth. We demonstrate that surface display of an enzyme from sea sponges, silicatein α, on the hyphae of the filamentous fungus Aspergillus niger enables mineralization of polysilicate and that this does not impact fungal growth. We also show that this strategy can be extended to other silicatein α variants and characterize how the degree of mineralization can be modulated. We then demonstrate that mineralization enhances the mechanical properties of the mycelium including its tensile strength, modulus, and toughness. Finally, we show how these reinforced mycelia can be grown without external carbon sources using a synthetic lichen-based coculture to facilitate low-cost biomanufacturing. Together, our results lay the groundwork for the sustainable production of mineralized mycomaterials and create a new model system to study how mineralization impacts growth and mechanical properties.

Phage genome engineering methods accelerate the study of phage biology, the discovery of new functions, and the development of innovative genetic engineering tools. Here, we present QuickPhage, a rapid, technically accessible, precise, and cost-effective method for engineering Bacillus subtilis phages. Our approach uses CRISPR-Cas9 as a counter-selection system to isolate mutants of the model lytic siphovirus phage, SPP1. Efficient genome editing was achieved using homologous repair patches as short as 40 nucleotides, enabling streamlined patch construction and parallel engineering, resulting in highly accurate genome edits within a day. We applied QuickPhage to delete both essential and nonessential phage genes and to insert reporter genes. Protein production, such as GFP, was synthetically regulated using inducible systems without significantly affecting phage fitness, achieving induction levels of up to 400-fold. Time-series coinfection experiments with fluorescent protein expressing phages also revealed a highly efficient superinfection arrest mechanism that prevents reinfection as early as 13 min after initial infection. These findings highlight the potential of phages for protein production, opening new opportunities for metabolic engineering. This work also lays the foundation for systematic phage genome refactoring workflows and the development of phage-based tools for efficient DNA delivery, thereby expanding the synthetic biology toolbox for B. subtilis.

Temporal transcriptional modulation of immune-related genes offers powerful therapeutic potential for treating inflammatory diseases. Here we introduce an enhanced zinc finger (ZF)-based transcriptional repressor delivered via lipid nanoparticles for controlling immune signaling pathways in vivo. By targeting Myd88, an essential adaptor molecule involved in immunity, our system demonstrates therapeutic efficacy against septicemia in C57BL/6J mice and improves repeated AAV administration by reducing antibody responses. This epigenetic engineering approach provides a platform for safe and efficient immunomodulation applicable across diseases caused by imbalanced inflammatory responses.