Validation study of 24 deepwell microtiterplates to screen libraries of strains in metabolic engineering

In this study we validated the use of 24 square deepwell microtiterplates to screen large libraries of metabolically engineered strains by investigating the optimization of succinate production. Wild type E. coli MG1655 and 11 derived mutants were physiologically evaluated by growth in 24 deepwell MTPs and 2L benchtop bioreactors. Growth parameters, product yields and byproduct formation were determined for all mutants. The results show that similar average values and standard deviations for these parameters were obtained. Especially a high correlation was noticed for the acetate byproduct yield and the succinate production rate. For these parameters there was no significant difference for 8 out of 12 strains between MTPs and 2L bioreactors. However a lower maximum growth rate was observed in 2L reactors as opposed to 24 deepwell plates for 9 out of 12 mutants which could be linked to the higher amount of dead cells in the benchtop bioreactors (12% vs. 2% in MTPs). Finally, a cluster-based approach was used to select good producer strains, i.e. strains with a high succinate yield and succinate production rate. Bad, intermediate and good producer strains were clustered in the same groups for MTPs and benchtop bioreactors for 11 out of the 12 investigated strains.