Integrating Machine Learning and Transcriptomics to Enhance β-Carotene Production in Yeast

ABSTRACT: Optimization of microbial production is essential for the application of synthetic biology in industrial and sustainable biosynthesis. β-carotene is a high-value compound that can be heterologously produced in budding yeast Saccharomyces cerevisiae, providing an alternative to natural extraction. However, intelligent design of strain and process to reduce the production cost remains a great challenge. In this study, we aimed to enhance β-carotene production by integrating systems biology and machine learning-guided strain design of the β-carotene-producing yeast. We identified transcriptional alterations in the β-carotene-producing yeast including genes associated iron deficiency, while no significant changes in the mevalonate (MVA) pathway. We then fine-tuned gene expression of rate-limiting enzymes in the mevalonate pathway through combinatorial construction of promoters and terminators. XGBoost training was applied in the DBTL cycle to facilitate rapid optimization. In the second DBTL cycle, fine-tuning MVA gene expression resulted in a 139% improvement in β-carotene titer. We then supplemented β-carotene production with iron, guided by transcriptional insights in alternations of genes related to iron uptake in β-carotene-producing yeast, resulting in an 70.54% improvement in β-carotene titer at 48 hours. Moreover, integrating the fine-tuned MVA cassette with iron supplementation yielded up to 72.07 mg/L of β-carotene at 72 hours, representing a 67.79% increase compared to that of the parent strain without MVA gene adjustments and iron supplementation. Our study highlights the potential of combining machine learning and omics approaches with synthetic biology to enhance non-native biochemical production in yeast.

ORGANISM(S): Saccharomyces cerevisiae

PROVIDER: GSE288397 | GEO | 2026/08/03

REPOSITORIES: GEO