Pre-mRNA splicing is a core process in eukaryotic gene expression, and splicing dysregulation has been linked to various diseases. However, very few small molecules have been discovered that can modulate spliced mRNA formation or inhibit the splicing machinery itself. This study presents a novel high-throughput screening (HTS) platform for identifying compounds that modulate splicing. Our platform comprises a two-tiered screening approach: A primary screen measuring growth inhibition in sensitized Saccharomyces cerevisiae (yeast) strains and a secondary screen that relies on production of a fluorescent protein as a readout for splicing inhibition. Using this approach, we identified 4 small molecules that cause accumulation of unspliced pre-mRNA in vivo in yeast. In addition, cancer cells expressing a myelodysplastic syndrome-associated splicing factor mutation (SRSF2P95H) are more sensitive to one of these compounds than those expressing the wild-type version of the protein. Transcriptome analyses showed that this compound causes widespread changes in gene expression in sensitive SRSF2P95H-expressing cells. Our results demonstrate the utility of using a yeast-based HTS to identify compounds capable of changing pre-mRNA splicing outcomes.

High-throughput microfluidics has transformed biomedical research by enabling precise and parallel sample handling, but most devices are single-use due to channel occlusion and contamination from experiments. Alongside low fabrication yield and reduced experimental success associated with dense microfeatures, this creates a major bottleneck for scalable high-throughput applications. We present a rapid, reusable, and modular high-throughput microfluidic platform with integrated microvalves for automation. The platform employs a multilayer architecture consisting of a custom casing, PDMS layers with dense microfeatures for fluid handling and culture, and a glass substrate. Permanent bonding is applied only between control and fluid layers, while reversible bonding is used at all other interfaces, including the substrate. Because substrate is the primary cell-contact surface and can be readily detached, the remaining layers can be disassembled, thoroughly cleaned, and reused with minimal processing on a new substrate. This approach improves repeatability and experimental success while reducing preparation time from days to ~2 hours. The disassemblable design also supports incorporation of application-specific layers between fluid layer and substrate, enhancing platform versatility for 3D culture. We validated performance through pressure/flow characterization and on-chip cell/organoid culture. Overall, our platform accelerates rapid high-throughput data generation across diverse biological applications.

Bacterial chromosomes contain Genomic Islands carrying genes involved in virulence, mutualism, and resistance to antibiotics that are transferred horizontally. Some of them conjugate autonomously (Integrative Conjugative Elements, ICEs), while others use relaxases to hitch on the former's conjugation machinery (Integrative Mobilizable Elements, IMEs). Yet, the mobility mechanism of most Genomic Islands remains elusive. Here, we explore the hypothesis that many carry origins of transfer (oriTs) by conjugation. Since very few known oriTs were found in ICEs and IMEs, we identified 52 novel families of oriTs that cover most integrative elements in six major nosocomial species. Most of these oriTs are specific to integrative elements, suggesting a clear split in the targets of hitchers of conjugative elements between plasmids and integrative elements. Among 7,363 Genomic Islands, we discovered more than 1,500 IMEs carrying an oriT and lacking relaxases and MPF systems. These elements, coined iOriTs, form diverse large ancient families that can be found in different species. These hitchers are genetically unrelated to the putative helpers, beyond the similarity at the oriT sequence, and often outnumber them. They include well-known pathogenicity islands, for which the mechanisms of mobility remained unknown. Unlike plasmids, iOriTs have few antibiotic resistance genes, but high density of virulence factors. Like plasmids, the vicinity of oriTs concentrates defense and counter-defense systems potentially favoring Genomic Islands dissemination. Hence, iOriTs are frequent integrative mobile genetic elements that evolved to transfer horizontally by hitching on other elements while contributing to bacterial genetic adaptation.

Genome editing enzymes have vast therapeutic potential. However, achieving sufficient delivery in vivo remains a major challenge, because editing machinery is confined to the subset of transfectable cells in a tissue. Here, we tested the possibility that genome editing could be amplified in vivo by programming transfected cells to produce and transfer editing enzymes in lipid vesicles to neighboring cells. Our data show that this NANoparticle-Induced Transfer of Enzyme (NANITE) strategy tripled editing efficiency in cultured cells relative to non-spreading controls. Furthermore, a single intravenous injection of the NANITE plasmid into mice induced ~3-fold higher levels of liver editing at the Ttr locus relative to non-spreading controls, with corresponding reductions in serum transthyretin levels. Amplifying therapeutic enzymes in situ offers a nonviral and non-infectious strategy to overcome low delivery efficiencies and reduce effective dose requirements.

Bacterial cellulose (BC) is produced by diverse bacterial species, including members of the genus Komagataeibacter, and has emerged as a potential sustainable alternative to conventional materials such as plastics and leather. Its production by bacteria is rapid and easily scalable, and via genetic modifications or culturing the bacteria with other microbes, it is possible to rationally enhance the BC that is produced, for example by adding functional components. Here, we systematically evaluated BC functionalization strategies, including genetic engineering and co-culturing with other engineered microbes across the most widely used Komagataeibacter species: K. rhaeticus, K. xylinus, K. medellinensis and K. sucrofermentans. We established that all tested strains are amenable to DNA transformation and capable of expressing heterologous genes from identical genetic constructs. However, we identified species-specific differences in heterologous gene expression, cellulose production in varying environmental conditions and in the abilities of the strains to co-culture with Escherichia coli. Additionally, we demonstrated that all these cellulose-producing bacteria can establish functional symbiotic co-cultures with yeast. While inter-species variations in heterologous gene expression and co-culture dynamics are evident, BC modification through genetic engineering and co-culturing strategies remains achievable across all tested Komagataeibacter species. Our work provides a comparative framework to guide researchers in selecting optimal species based on their specific application requirements.

Advances in synthetic biology continue to potentiate bacterial cancer therapy. Here, we constructed biosensor-driven encapsulation systems for autonomous control of capsular polysaccharides of Escherichia coli Nissle 1917 to improve pharmacokinetic profiles. The engineered bacteria were programmed to express capsular polysaccharides for immune evasion upon intravenous administration to reach tumors and then turn off gene expression upon colonizing the tumors based on quorum-sensing or acid-sensing to prevent dissemination of bacteria into the systemic circulation. Because a classical pharmacokinetic model could not capture the dynamic nature of living therapeutics, a two-state pharmacokinetic model was developed to simulate the autonomous control of capsular polysaccharides in different biological compartments and their impact on biodistribution. Using this model, we identified parameters in gene circuit dynamics and immune clearance that influence tumor colonization and systemic bacterial persistence. In a “humanized” pharmacokinetic model with an increased rate of complement-mediated lysis of bacteria, biosensor-driven systems achieved tumor seeding densities comparable to wild-type bacteria while reducing bacterial loads in blood and liver by several orders of magnitude, highlighting their potential for safe systemic delivery. The biosensor-driven systems represent a more effective strategy to control living drugs than inducible systems, and the two-state pharmacokinetic model is a first step to capture the autonomous nature of this new class of therapeutics for clinical translation.