Actinomycetes are important producers of pharmaceuticals and industrial enzymes. However, wild type strains require laborious development prior to industrial usage. Here we present a generally applicable reporter-guided metabolic engineering tool based on random mutagenesis, selective pressure, and single-cell sorting. We developed fluorescence-activated cell sorting (FACS) methodology capable of reproducibly identifying high-performing individual cells from a mutant population directly from liquid cultures. Actinomycetes are an important source of catabolic enzymes, where product yields determine industrial viability. We demonstrate 5-fold yield improvement with an industrial cholesterol oxidase ChoD producer Streptomyces lavendulae to 20.4 U g⁻¹ in three rounds. Strain development is traditionally followed by production medium optimization, which is a time-consuming multi-parameter problem that may require hard to source ingredients. Ultra-high throughput screening allowed us to circumvent medium optimization and we identified high ChoD yield production strains directly from mutant libraries grown under preset culture conditions. Genome-mining based drug discovery is a promising source of bioactive compounds, which is complicated by the observation that target metabolic pathways may be silent under laboratory conditions. We demonstrate our technology for drug discovery by activating a silent mutaxanthene metabolic pathway in Amycolatopsis. We apply the method for industrial strain development and increase mutaxanthene yields 9-fold to 99 mg l⁻¹ in a second round of mutant selection. In summary, the ability to screen tens of millions of mutants in a single cell format offers broad applicability for metabolic engineering of actinomycetes for activation of silent metabolic pathways and to increase yields of proteins and natural products.

放线菌是药物和工业酶的重要生产者。然而，野生型菌株在工业使用之前需要费力的开发。在这里，我们提出了一种基于随机诱变、选择压力和单细胞分选的普遍适用的报告指导代谢工程工具。我们开发了荧光激活细胞分选 (FACS) 方法，该方法能够直接从液体培养物中重复识别来自突变群体的高性能单个细胞。放线菌是分解代谢酶的重要来源，其产品产量决定了工业可行性。我们展示了工业胆固醇氧化酶 ChoD 生产商Streptomyces lavendulae的 5 倍产量提高至 20.4 U g ^-1在三轮中。传统上，菌株开发之后是生产培养基优化，这是一个耗时的多参数问题，可能需要难以获取成分。超高通量筛选使我们能够规避培养基优化，我们直接从在预设培养条件下生长的突变文库中鉴定出高产量的 ChoD 菌株。基于基因组挖掘的药物发现是一种很有前途的生物活性化合物来源，但由于观察到目标代谢途径在实验室条件下可能是沉默的，这一点变得复杂。我们通过激活Amycolatopsis中的沉默的mutaxanthene 代谢途径来展示我们的药物发现技术。我们应用该方法进行工业菌株开发，将互蒽醌产量提高 9 倍至 99 mg l^-1在第二轮突变体选择中。总之，以单细胞形式筛选数千万个突变体的能力为放线菌的代谢工程提供了广泛的适用性，以激活沉默的代谢途径并提高蛋白质和天然产物的产量。