Abstract:[Background] Xiamenmycin A, which shows significant anti-fibrotic activity and can serve as a potential medicinal candidate, is a major secondary metabolite of Streptomyces xiamenensis 318. However, the yield of xiamenmycin A in the wild-type strain is only 14 mg/L. Thus, it is urgent to improve the production of this compound in the strain. [Objective] Random mutagenesis and resistance reporter-based selection was used to develop a high xiamenmycin A-producing strain and further improve the yield of the compound through medium optimization. [Methods] An antibiotic resistance gene was fused at the 3' end of the last gene (ximE) of xiamenmycin A biosynthesis gene cluster, and thus the expression of the entire gene cluster could be indicated by antibiotic resistance level. After atmospheric and room temperature plasma (ARTP) treatment, the high-yielding strain was selected as the mutant with high antibiotic resistance level. The yield of xiamenmycin A in the mutant was further enhanced by medium optimization. [Results] A neo -labeled strain MT-XN was constructed as a starting strain. After one round of ARTP mutagenesis, a mutant MA-8 with resistance to 90 mg/L kanamycin and xiamenmycin A yield of 101.7 mg/L was obtained. Moreover, the yield of xiamenmycin A in MA-8 reached 134.2 mg/L with the medium optimized by response surface methodology, 845.1% higher than that in the wild-type strain. [Conclusion] Random mutagenesis combined with reporter-based selection can be used to identify high xiamenmycin A-yielding strain, which lays a foundation for drug development of this compound.