In vivo continuous evolution via phenotypic sorting to alleviate metabolic bottlenecks in β-alanine production

β-Alanine is an important platform chemical whose biosynthesis efficiency is limited by pathway bottlenecks and enzyme constraints. In this study, to overcome genetic interactions for β-alanine production, we systematically engineered Escherichia coli MG1655 through modular pathway optimization and combinatorial regulation. To overcome the limitations of L-aspartate-α-decarboxylase from Bacillus subtilis (PanD_bsu), we developed an in vivo evolution platform combining base-editing systems with biosensor guidance, thus generating PanD_bsu variants with enhanced activity. This system facilitated high-throughput screening and real-time monitoring of β-alanine production and accelerated mutant selection. Furthermore, site saturation and iterative mutations identified a beneficial PanD_bsu^T4E mutant, which enhanced specific β-alanine production in engineered strain MA31 by 62.45%. Structural and functional analysis revealed that PanD_bsu^T4E stabilized its quaternary structure via a Glu-Lys salt bridge. This work describes a scalable strategy for addressing pathway bottlenecks and highlighted the potential of integrating protein engineering with biosensor-guided evolution to optimize microbial cell factories.