This technology uses machine learning to engineer improved enzymes that dramatically boost gene expression in eukaryotic cells, enabling more efficient production of mRNA, proteins, and therapeutics for research, industrial, and medical applications.
Background
The field of eukaryotic gene expression is foundational to biotechnology, synthetic biology, and therapeutic development, as it underpins the ability to produce proteins, enzymes, and mRNA-based therapeutics in living cells. Efficient gene expression in eukaryotic systems is essential for applications ranging from metabolic engineering and industrial biomanufacturing to mRNA vaccine and therapeutic protein production. However, achieving high levels of gene expression in eukaryotes is inherently challenging due to the complexity of their cellular machinery, including the need for precise mRNA modifications such as capping, which are critical for mRNA stability, nuclear export, and translation. As demand grows for scalable, high-yield, and reliable gene expression systems in research and industry, there is a pressing need for technologies that can overcome the intrinsic limitations of current enzymatic tools.
Current approaches to enhancing gene expression in eukaryotic systems often rely on wild-type bacteriophage-derived enzymes, such as T7 RNA polymerase and viral capping enzymes, which are optimized for prokaryotic contexts and do not efficiently accommodate the sophisticated post-transcriptional modifications required in eukaryotes. Traditional enzyme engineering strategies—such as random mutagenesis and directed evolution—are labor-intensive, time-consuming, and typically limited in their ability to explore the full mutational landscape, especially outside enzyme active sites. These methods frequently result in incremental improvements and may overlook synergistic mutations that could dramatically boost enzyme performance. Moreover, existing systems tend to address only isolated steps in the gene expression pathway, failing to resolve multiple bottlenecks simultaneously, which limits overall gains in mRNA yield, stability, and translation efficiency. As a result, current gene expression platforms often suffer from suboptimal protein output, poor mRNA stability, and high production costs, constraining their applicability in advanced therapeutic and industrial settings.
Technology Description
This technology is a machine learning-guided platform for engineering T7 RNA polymerase and mRNA capping enzymes to significantly boost gene expression efficiency in eukaryotic systems. It leverages advanced neural network models—such as MutCompute, MutComputeX, MutRank, and Stability Oracle—to predict beneficial mutations across the entire enzyme structures, not just at active sites. These predictions are experimentally validated in yeast using a high-throughput fluorescence-based screening system, with iterative combination of advantageous mutations to maximize performance. The result is a suite of engineered enzyme variants, including EvoT7 (a sextuple mutant T7 RNA polymerase) and EvoBMCE (a triple mutant capping enzyme), as well as fusion proteins that link these enzymes for synergistic effects. The platform enables up to 12.68-fold increases in gene expression by simultaneously enhancing transcriptional activity and mRNA stability, with modular designs suitable for diverse biological applications.
What differentiates this technology is its comprehensive, synergistic approach to overcoming multiple bottlenecks in eukaryotic gene expression. Unlike traditional methods that rely on random mutagenesis or focus on optimizing single enzymatic steps, this platform uses machine learning to systematically explore vast mutational landscapes, including residues distant from active sites, and to predict combinations of mutations with additive or synergistic effects. The integration of ML-guided engineering with scalable yeast-based validation accelerates the discovery process, reduces experimental workload, and uncovers beneficial mutations that conventional approaches might overlook. Its modular fusion protein architecture, ability to enhance both transcription and post-transcriptional processes, and broad applicability—from synthetic biology and therapeutics to industrial biomanufacturing—set it apart as a transformative solution for next-generation gene expression technologies.
Benefits
- Significantly enhances gene expression efficiency in eukaryotic systems by improving both transcriptional activity and mRNA capping.
- Employs advanced machine learning models to precisely predict beneficial enzyme mutations, enabling faster and more effective enzyme optimization than traditional methods.
- Engineered fusion proteins synergistically overcome transcriptional and post-transcriptional bottlenecks, achieving up to 12.68-fold increase in gene expression.
- Improves enzyme stability, substrate specificity, and thermostability, resulting in higher RNA yield and protein expression.
- Modular design allows flexible application across diverse biological contexts, including research, therapeutic, and industrial uses.
- Validated in scalable yeast-based systems, facilitating cost-effective, high-throughput screening and bulk RNA production.
- Potential to reduce production costs and improve accessibility of mRNA vaccines, therapeutics, and biomanufacturing processes.
Commercial Applications
- mRNA vaccine manufacturing
- Therapeutic protein production
- Cell-free transcription/translation systems
- Metabolic engineering for biomanufacturing
- CRISPR guide RNA synthesis
Additional Information
This technology employs machine learning to engineer T7 RNA polymerase (EvoT7) and mRNA capping enzymes (EvoBMCE). EvoT7 boosts transcription 6.16-fold, while EvoBMCE improves capping 3-fold. Their fusion synergistically enhances eukaryotic gene expression up to 12.68-fold, addressing transcriptional and post-transcriptional bottlenecks for improved mRNA production and stability.
Patent PCT/US2026/020360 filed 03/23/26
Publication
Machine learning-guided engineering of T7 RNA polymerase and mRNA capping enzymes for enhanced gene expression in eukaryotic systems
