Optimization and modelling of protein refolding conditions

Bernd Anselment, doctoral thesis Technische Universität München, 2012

A bottleneck for the production of recombinant proteins in Escherichia coli is often the limited in vivo solubility. Refolding of the target protein becomes necessary and complicates downstream processing. Individual refolding conditions have to be determined empirically with a large experimental effort. Thus, a novel stochastic optimization strategy was developed on the basis of database information. This strategy allows an optimization of protein refolding conditions with a minimum number of experiments. The suitability of this new approach was demonstrated with six different proteins achieving an up to 30-fold increase of refolded activities compared to literature. In addition, it was shown that bagged decision trees were able to model the refolding conditions provided that the dataset of the optimization was large enough.

Publications

• Anselment B, Schoemig V, Kesten C, Weuster-Botz D (2012): Statistical versus stochastic experimental design – an experimental comparison on the example of protein refolding. Biotechnol Prog 28: 1499-1506.
• Anselment B, Baerend D, Mey E, Buchner J, Weuster-Botz D, Haslbeck M (2010): Experimental optimization of protein refolding with a genetic algorithm. Protein Science 19: 2085-2095.