Adaptive cargo deletion as an evolutionary dead-end during phage-based directed evolution

The purpose of this project was to evaluate the performance and reliability of the P1 phage-based inducible directed evolution (IDE) method for performing global transcriptional machinery engineering (gTME) to improve pH tolerance in Escherichia coli Nissle. Unlike prior gTME approaches that evolved single sigma factors, we simultaneously mutated three sigma factors associated with pH tolerance (rpoD, rpoN, and rpoH) through five rounds of mutagenesis and growth-based selection at pH 4.2.

Mutations in rpoD (V98E and R603C) emerged during early rounds and reduced the lag phase under acidic conditions. However, library improvement plateaued after the third cycle. In later rounds, an empty vector resulting from the deletion of plasmid-encoded sigma factors became dominant. This deletion of plasmid-encoded sigma factors improved low-pH fitness, but it prevented further mutations and led to an evolutionary dead end. Additionally, transcriptome analysis showed that the single mutations (V98E, R603C) that preceded sigma factor deletion improved fitness through mechanisms resembling those observed upon sigma factor deletion. Together, these findings indicate that phage-based and continuous directed evolution approaches may face challenges if the evolutionary substrate itself reduces performance relative to a wild-type strain.