Chapter 3 High-Throughput Expression Screening and Purification of Recombinant Proteins in E. coli Natalie J. Saez and Renaud Vincentelli Abstract The protocols outlined in this chapter allow for the small-scale test expression of a single or multiple proteins concurrently using several expression conditions to identify optimal strategies for producing soluble, stable proteins. The protocols can be performed manually without the need for specialized equipment, or can be translated to robotic platforms. The high-throughput protocols begin with transformation in a 96-well format, followed by small-scale test expression using auto-induction medium in a 24-well format, finishing with purification in a 96-well format. Even from such a small scale, there is the potential to use the purified proteins for characterization in pilot studies, for sensitive micro-assays, or for the quick detection of and differentiation of the expected size and oxidation state of the protein by mass spectrometry. Key words E. coli, Bacteria, Expression, Recombinant, High-throughput, Purification, Autoinduction, Immobilized metal affinity chromatography (IMAC), TEV cleavage
1
Introduction Traditionally, protein production approaches have centered on the case-by-case exploration of proteins of particular interest. With advances in genomics and thousands of novel and interesting proteins being discovered at such an accelerated rate, these production strategies have become outdated, causing a bottleneck in structural and functional studies. Parallelization of these traditional approaches into high-throughput pipelines at a small scale allows the screening for optimal expression conditions, enabling the testing of various parameters on soluble expression levels. This may include, but is not limited to, using varying expression strains [1, 2], temperature [3, 4], media [2, 3], target variants [5], fusion partners [6–12], coexpression with chaperones [13, 14], cytoplasmic or periplasmic expression [15], and purification buffer components [3]. Testing all of these variables using traditional methods would be highly inefficient. However, by implementing high-throughput approaches,
Yu Wai Chen (ed.), Structural Genomics: General Applications, Methods in Molecular Biology, vol. 1091, DOI 10.1007/978-1-62703-691-7_3, © Springer Science+Business Media, LLC 2014
33
34
Natalie J. Saez and Renaud Vincentelli
up to 96 expressions and purifications can be tested in parallel, and multiple can be performed in any 1 week. This means that many variables can be tested on a few targets or a few variables can be tested on many targets with a high level of efficiency. The strategy also provides good reproducibility upon scale-up as the same culture and purification conditions are utilized at both stages. Our high-throughput strategy utilizes E. coli, taking advantage of its ease of use, fast growth rates, relatively low cost of production, and adaptability to scaling up cultures for large-scale expression once optimal conditions are identified. With the range of E. coli strains available, the system is also applicable to a wide range of targets, even those that are not codon optimized for expression in E. coli (using strains that compensate for rare codons) or for targets with complex folds, containing multiple disulfide bonds (using strains that modify the reducing environment of the cytoplasm). Given that a large proportion of constructs are generally cloned directly without codon optimization, for the protocol described herein we have chosen to utilize the Rosetta 2 (DE3) pLysS strain, which carries tRNAs for rare codons that are not highly expressed in E. coli. However, it is possible for the highthroughput protocol to be trialled using different strains to check for variances in the soluble expression levels of target proteins and to continue optimization on the most desirable strain. We also use auto-induction [16], which simplifies expression, eliminating the need for manual induction. One downfall of expression in E. coli is that it does not allow for posttranslational modifications, other than disulfide bonding, and for these types of targets alternative production methods need to be sought (either by expression in eukaryotes or by chemical synthesis for small proteins (
1
Introduction Traditionally, protein production approaches have centered on the case-by-case exploration of proteins of particular interest. With advances in genomics and thousands of novel and interesting proteins being discovered at such an accelerated rate, these production strategies have become outdated, causing a bottleneck in structural and functional studies. Parallelization of these traditional approaches into high-throughput pipelines at a small scale allows the screening for optimal expression conditions, enabling the testing of various parameters on soluble expression levels. This may include, but is not limited to, using varying expression strains [1, 2], temperature [3, 4], media [2, 3], target variants [5], fusion partners [6–12], coexpression with chaperones [13, 14], cytoplasmic or periplasmic expression [15], and purification buffer components [3]. Testing all of these variables using traditional methods would be highly inefficient. However, by implementing high-throughput approaches,
Yu Wai Chen (ed.), Structural Genomics: General Applications, Methods in Molecular Biology, vol. 1091, DOI 10.1007/978-1-62703-691-7_3, © Springer Science+Business Media, LLC 2014
33
34
Natalie J. Saez and Renaud Vincentelli
up to 96 expressions and purifications can be tested in parallel, and multiple can be performed in any 1 week. This means that many variables can be tested on a few targets or a few variables can be tested on many targets with a high level of efficiency. The strategy also provides good reproducibility upon scale-up as the same culture and purification conditions are utilized at both stages. Our high-throughput strategy utilizes E. coli, taking advantage of its ease of use, fast growth rates, relatively low cost of production, and adaptability to scaling up cultures for large-scale expression once optimal conditions are identified. With the range of E. coli strains available, the system is also applicable to a wide range of targets, even those that are not codon optimized for expression in E. coli (using strains that compensate for rare codons) or for targets with complex folds, containing multiple disulfide bonds (using strains that modify the reducing environment of the cytoplasm). Given that a large proportion of constructs are generally cloned directly without codon optimization, for the protocol described herein we have chosen to utilize the Rosetta 2 (DE3) pLysS strain, which carries tRNAs for rare codons that are not highly expressed in E. coli. However, it is possible for the highthroughput protocol to be trialled using different strains to check for variances in the soluble expression levels of target proteins and to continue optimization on the most desirable strain. We also use auto-induction [16], which simplifies expression, eliminating the need for manual induction. One downfall of expression in E. coli is that it does not allow for posttranslational modifications, other than disulfide bonding, and for these types of targets alternative production methods need to be sought (either by expression in eukaryotes or by chemical synthesis for small proteins (
