Design of Artificial Enzymes for the Baylis-Hillman reaction

The aim of this proposal is the design of an artificial enzyme, catalyzing a reaction not occurring in nature. Enzymes are proteins which work as biological catalysts enhancing the reaction rate of chemical reactions in biological systems. As target reaction the Baylis-Hillman reaction was chosen. During this reaction a new bond between two carbon atoms is formed. The derived products are used as drugs or as building blocks thereof. Of special interest are chiral molecules. These are molecules which are not superimposable, like the right hand and the left hand. For chemical synthesis of those molecules proteins can be used. They give as product just one of the two forms. Our goal is to synthesize with our method one the one hand a chiral building block of an antineoplastic drugs used in the treatment of cancer and one the other hand a chiral pheromone to attract an insect, the wheat weevil, one of the largest pests damaging stored grain. With our methods a more sustainable and environmental more friendly synthesis would be possible. For the design of such an enzyme different protein scaffolds will be chosen from literature and using computer programs. They will provide a pocket where a small organic molecule (organocatalyst) is mounted. This is the actual catalyst, while the protein provides a pocket, where just one of the mirror-imaged molecules can be synthesized. The general structure of the derived enzyme will be analyzed in detail to gain knowledge about the exact shape of the pocket. This information is applied to improve chiral synthesis as well as the activity. For this reason, different computer programs for protein design will be employed. There are several artificial enzymes described in literature, but almost all of them use metal- based catalysts. There is just one case reported where an organocatalyst was employed. For a Baylis-Hillman reaction such an approach was never undertaken. This new methodology combines now artificial enzymes with organocatalysts and computational design of those. New insights in the field of protein design as well as in the construction of artificial enzymes can be gained during this project. To facilitate this research, the Protein Design group under the lead of Prof. Dr. Höcker at the University of Bayreuth was chosen. Her experience in protein design and structure determination will complement the qualifications of the applicant, who has a strong background in biocatalysis in combination with organic synthesis and also experience in protein design software.

Enzymes are proteins with catalytic function. They are considered to be green catalysts, as they are produced by an organism. They often have an additional great advantage since they are very selective for certain reactions or substrates and are able to catalyze the formation of products which cannot gained with "classical" catalysts. The drawback is that nature provides us only enzymes for a limited set of reaction and substrates. The chemist might need different substrates and activities than found in nature for a certain reaction needed in industry, for example in a "greener" route for the production of pharmaceuticals. To gain new functions and to design enzymes several approaches are used. We were using some of them to design an enzyme able to catalyze a reaction not catalyzed by any natural enzyme. The targeted reaction facilitates carbon-carbon bond formation, a reaction of high interest, in a very elegant and efficient way. One of the aforementioned approaches to extent the reaction scope of enzyme is computational enzyme design. But this way of designing novel enzymatic activities is still troublesome due to low initial activities if compared to the ones reported from enzymes evolved by nature. We are not able to model and describe all properties of a new enzymes such that it can be calculated by a computer in a reasonable timescales, we rely on simplified models. One strategy to overcome this problem of initial activity is the use of a so-called cofactor bearing already the desired catalytic functionality. Cofactors are used by nature as well - some of the vitamins are a cofactor. They are catalysts able to catalyze a reaction especially well when embedded in a protein. We used a variation of an already described catalyst usually not used in nature and embedded it into a protein. The reaction was computationally modeled and variants of protein were made to position and accommodate the catalyst as well as the substrates. This problem is less complex than designing a totally new enzyme since the catalytic activity is already present. After several round of protein structure determination and design a new proteinogenic catalyst was gained carrying out the targeted reaction. We were able to synthesize the cofactor, embed it into a redesign protein, solve the structure of the catalyst and do some theoretical work on the mechanism as well as to synthesize with the new catalyst our model target.