A standardized toolkit for programmable protein secretion and surface display in Bacteroides thetaiotaomicron
Yeh, Y.-H.; Sirk, S. J.
Show abstract
Bacteroides thetaiotaomicron (Bt), a dominant bacterial species in the human gut, is a promising chassis for engineering in situ therapeutic delivery systems. Previously, we developed a secretion toolkit composed of endogenous lipoprotein signal peptides (SPs) and full-length secretory proteins from Bt. However, due to variations in length, structure, and amino-acid sequence, these SPs exhibited inconsistent secretion efficiencies across different cargo proteins. Because the activity of individual SP-cargo pairs is not readily predictable, screening and optimization is often required to achieve target secretion titers. To enhance the utility of our toolbox, we studied the impact of different SP sequence components on protein secretion, then applied this knowledge to develop a standardized toolkit to and enable predictable and tunable protein secretion across diverse SP-cargo pairs. To achieve this, we first identified the lipoprotein export sequence (LES) as the key determinant of efficient secretion of heterologous proteins by lipoprotein SPs. We next performed mutagenesis on the LES region of a representative lipoprotein SP to generate a pool of mutants featuring a standardized SP backbone with diversified LES regions. Screening and characterization of this mutant pool revealed a charge-dependent regulation of both secretion and surface display of heterologous cargo proteins. From these findings, we established a toolkit with improved tunability, enhanced predictability, and surface display capabilities that minimizes the need for iterative screening when developing protein secreting gene circuits for Bt and other Bacteroides species. By enhancing both the flexibility and control of therapeutic protein output, these results expand the potential of engineered living therapeutic applications, particularly those requiring tunable dosing or surface presentation of proteins.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.