Biocatalytic transamination with recombinant Saccharomyces cerevisiae: Challenges and opportunities

Sammanfattning: Chiral building blocks are important molecules for synthesis of fine chemicals and pharmaceuticals. Biocatalysis has gained in relevance over traditional organic methods for synthesis of chiral compounds, as reactions can be performed at high enantio-selectivity and purity with environmentally friendly, simple, and cheap methods. The aim of the present study was to explore the feasibility of using metabolically active microorganisms as whole-cell biocatalysts for the production of building blocks containing chiral amine and alcohol functionalities, using both strain and process engineering. I also discuss the possibilities and challenges of finding new enzymes for biocatalytic processes by the construction and screening of metagenome libraries. This thesis describes the first exploitation of the metabolism of Saccharomyces cerevisiae for the generation of chiral amines via transamination. As a model transamination reaction, the kinetic resolution of racemic 1-phenylethylamine (PEA) to (R)-1-phenylethylamine ((R)-1-PEA) and acetophenone (ACP) was used. A novel ω-transaminase from the plant Capsicum chinense was first evaluated in vitro and was shown to be active at a broad range of pH and to use amine acceptors that can be derived from metabolism of glucose in S. cerevisiae. Saccharomyces cerevisiae and Escherichia coli were then compared as hosts for the transaminase from C. chinense, with pyruvate as amine acceptor. Recombinant S. cerevisiae showed lower initial reaction rates than recombinant E. coli. However, yeast had a significantly higher degree of robustness and reached a much higher enantiomeric excess (ee) than E. coli. Transamination was also achieved with glucose as co-substrate in recombinant S. cerevisiae, as pyruvate could be derived from glycolysis. It was also possible to omit the co-factor PLP in the transamination reaction, demonstrating that even PLP could be derived from the metabolism. Comparison of pyruvate and glucose as co-substrate showed that conversion of racemic 1-PEA was higher with glucose. In addition, the viability of the cells was significantly higher with glucose than with pyruvate. Two chiral compounds, an amine and an alcohol, were then synthesized in a cascade reaction system. The applied model reaction was racemic 1-PEA to (R)-1-PEA and (R)-1-phenylethanol ((R)-1-PE) with co-expression of a transaminase from Capsicum chinense and a reductase from Lactobacillus kefir. The reduction of ACP, which is the by-product of the first reaction, led to decreased inhibition of the transamination reaction and thus to a higher overall conversion of the racemic substrate. The reduction was very efficient, and an ee of > 99% was achieved. A three-fold increase in transaminase gene copy number also led to a two-fold increased conversion, revealing the transamination as a controlling step. Comparison of resting and growing cells showed that even if the growing cells did not result in higher overall conversion of the racemic substrate, the reduction was improved. This was most probably due to better regeneration of the co-factor NADPH, demonstrated by lower accumulation of ACP and more formation of by-product from glucose metabolism under these conditions. The importance of the enzyme selection was demonstrated by comparing three transaminases from C. chinense, Chromobacterium violaceum, and Ochrobactrum anthropi in S. cerevisiae. It was found that the transaminase from C. violaceum had the highest reaction rate and conversion efficiency of the model reaction (racemic 1-PEA to (R)-1-PEA). This gene was expressed in an engineered strain that should accumulate pyruvate, showing slightly higher conversion than the control strain. In a parallel approach, omission of thiamine in the cultivation medium led to significantly higher conversion if the co-factor PLP was also omitted. Complete kinetic resolution of racemic 1-PEA was achieved with a strain containing several copies of C. violaceum transaminase―and with higher cell loading, no PLP, and glucose as sole co-substrate. Overall, the work described in this thesis is a step towards a new whole-cell biocatalysis process concept where the microorganism is exploited and engineered as a microbial cell factory for the generation of fine chemicals. In particular, it has demonstrated that sugar metabolism can be redirected towards inexpensive supply of co-substrates and co-factors.