nonheme Fe(II) hydroxylases: from structure-function relationships to redesign

Selective oxygenations of non-activated carbon atoms by O2 are a major target in green chemistry, as they are typically inaccessible by abiotic synthetic means. Nature, by contrast has evolved structures that allow the highly selective oxidation of organic molecules by O2 , the two most versatile being the P450 heme and the nonheme Fe(II) centers. Despite their unique catalytic potential, and contrary to the P450s, nonheme Fe(II) enzymes are not generally used as a platform for designed oxygenations. Reasons therefore may include their instability outside the cell and complications regarding the establishment of reliable, efficient whole cell screening methods for these systems. These factors limit the applicability of evolutionary methods for enzyme engineering. Rational methods of protein redesign that are rooted in an in depth understanding of the interplay of enzyme structure and function may help to bridge this gap. a-KG dependent nonheme Fe(II) hydroxylases (a-KG-MNH) are an enzyme grouping that transforms cell metabolites via oxidative hydroxylation. They all share a cupin fold and a common metal center organization. Yet, particular exponents of the a-KG-MNHs show high but distinct stereo-, regio-, and substrate-selectivities. In this project the structural basis of this diversity, which is not well understood, will be explored. Therefore exponents of a-KG-MNHs are subjected to a combination of experimental and computational methods. Mutational analysis and kinetic and spectral characterizations are correlated with molecular dynamic studies in order to gain insights into the impact of the protein structure and on particular steps of catalysis. The goal of the study is to introduce new catalytic functions into the protein scaffolds. The applicability of molecular dynamic studies to rationalize and predict catalytic properties in silico based on quantitative descriptors will be assessed. Results will furthermore be put in context with available data of a-KG-MNHs from literature and sequence data, in order to define the structural motifs that bring about stereo-, regio-, and substrate-selectivity in a-KG-MNHs. The overriding aim of the project is to increase the knowledge basis regarding the interplay of protein structure and catalytic properties of a-KG-MNH, in order to facilitate their computer assisted redesign and to expand their catalytic repertoire. Ultimately, this may open new routes for the sustainable biosynthetic production of added-value chemicals from feed stock.

The objective of this project was to understand on a molecular level how nonheme iron(II) bearing proteins burn organic molecules in a controlled, selective manner. With this knowledge, tools can be developed to design smart burners that catalyze new, tailor made reactions. The burning of organic material is generally considered a chaotic process producing CO2 and grime. Even for synthetic chemists the controlled use of molecular oxygen (O 2) for controlled oxidations is mostly still elusive. Nature on the other hand has evolved systems that are able to use O2 in the controlled synthesis of molecules: Nonheme-iron(II)- oxygenases (NHIOs) are proteins that are masters in burning organic molecules in a smart way by inserting oxygen molecules very selectively into the target molecules, reactions for which they have been tailored by evolution. This has been achieved by enveloping an iron(II) center, which activates O2 with a protein structure, which directs and positions the respective organic target molecule in a way that the activated oxygen atom can very selectively attack and then be incorporated into a specific position of the target molecule. The precise mechanisms by which the protein structure directs the target molecule are generally still not well understood. An objective of this project was to elucidate, how protein structures enable this selective process. To do this, the interactions of selected protein structures and their interaction with organic target molecules were computationally studied, and the structural features responsible for directing the molecules respective target positions for oxygenation to the iron center were identified. Consequently, the respective structures were changed via mutational analysis, and the catalytic activities were characterized in vitro and compared to computational simulations in an iterative process. The computational process was thus refined and automated and resulted in a computational platform that in principle facilitates the prediction and design of NHIOs with altered O2 insertion patterns. In parallel, a system has been established for the high-throughput screening of NHIOs. The resulting system has been used to design NHIOs that produce enantiopure (R)-madelic acid (note that natural enzymes can only perform the mirror reaction and produce (S)-mandelic acid) and NHIOs with altered amino acid hydroxylation patterns. In summary the project has created a platform that can help design biocatalysts which open new routes for the sustainable, fermentative production of pharmaceutical building blocks.