Screening factors effecting a response in soluble protein expression: Formalized approach using design of experiments

Abstract

We have integrated high-throughput expression and purification with quantitative protein analysis to identify factors influencing protein production. Application of high-throughput expression and purification, combined with automated gel capillary electrophoresis, allowed the quantitative analysis of multiple expression variables in a single experiment. An experimental design utilizing multiple factorial screens was employed to identify single factors and interactions having a significant impact on expression. As a test case, expression of the nonstructural protein NS3 from different hepatitis C virus genotypes (1b, 2a, and 3a) was examined in Escherichia coli. The 1b genotype of NS3 produced the highest level of expression, which was then further optimized using response surface modeling to give a four-fold increase in soluble protein levels. The quantitative and statistical approach presented has the capability of rapidly identifying interactions among experimental variables, leading to more reliable prediction of protein expression. We propose that this technique has universal application in the production of recombinant proteins, providing a powerful tool for the optimization of protein expression.