Effects of medium composition and nutrient limitation on loss of the recombinant plasmid pLG669-z and β -galactosidase expression by Saccharomyces cerevisiae

The effects of medium composition, nutrient limitation and dilution rate on the loss of the recombinant plasmid pLG669-z and plasmid-borne β -galactosidase expression were studied in batch and chemostat cultures of Saccharomyces cerevisiae strain CGpLG. The difference in growth rates between plasmid-free and plasmid-containing cells (Δμ) and the rate of segregation (R) were determined and some common factors resulting from the effect of medium composition on plasmid loss were identified. Glucose-limited chemostat cultures of CGpLG grown on defined medium were more stable at higher dilution rates and exhibited Δμ -dominated plasmid loss kinetics. Similar cultures grown on complex medium were more stable at lower dilution rates and exhibited R-dominated plasmid loss kinetics. Overall plasmid stability was greatest in phosphate-limited chemostat cultures grown on defined medium and was least stable in magnesium-limited cultures grown on defined medium. Δμ decreased and R increased with increased dilution rate, irrespective of medium composition. Increased plasmid loss rates at high or low dilution rates would appear to be characteristic of loss kinetics dominated by R or Δμ, respectively. Growth of glucose-limited chemostat cultures on complex medium decreased Δμ values but increased R values, in comparison to those cultures grown on defined medium. Any increased stability that a complex medium-induced reduction of Δμ may have conferred was counteracted by an increased R value. Increased β-galactosidase productivity was correlated with increased plasmid stability only in glucose-limited chemostat cultures grown on defined medium and not in those grown on complex medium. Previous studies have yielded contrasting responses with regard to the effect of dilution rate on recombinant plasmid loss from S. cerevisiae. Our findings can account for these differences and may be generally valid for the stability of similar yeast plasmid constructs. This information would facilitate the design of bioprocesses, where recombinant plasmid instability results in reduced culture productivity.