Resumen de Computational design of artificial metalloenzymes
Enzyme design is the scientific field that aims at discovering and/or optimizing biomolecules to reach new-to-Nature reactions. It is an area in wild expansion and constitutes one of the cornerstones of the transition of chemical practices towards greener alternatives.
An elegant way to construct novel biocatalysts is through the embedding of organometallic cofactors into biological scaffolds, leading to the so-called Artificial Metalloenzymes (ArMs). These hybrids bridge the catalytic versatility of the organometallic compound with the substrate and spatial specificity of the biological host. The design of ArMs has spread increasingly during the last two decades taking clear advantage of the major expansion of structural biology and the maturity of organometallic catalysis. Molecular modelling aims to help designers to provide with structural information that could serve for constructing optimum biocatalysts. However, despite the increasing improvement of the computation performance and the exponential development of new simulation techniques, the complexity of dealing with transition metal including systems has promoted modellers not to explore the ArM constructs.
The InSiliChem group, in which this Ph.D. has been performed, has focused on developing a specific framework for the study and design of ArMs. In particular, this has been based on the development of in silico multiscale strategies including standard computational methods.
This Ph.D. aims at increasing the potentiality of our computational platform for ArM design by 1) including classical Molecular Dynamics simulations into the integrative computational framework and 2) testing the validity of the methodology for the design of real case ArMs.
The results obtained could be summarize as follows: • The catalytic mechanism of two novel ArMs were decoded using the updated computational pipeline. These were a copper-Phenanthroline containing hydratase based on the Lactococcus Multidrug Resistance Regulator (LmrR) and a variety of novel mutants based on Streptavidine-Noyori complexes for cyclic imine reduction reaction. The study revealed the importance of the contribution of the MD simulations to decode the catalytic mechanism of these ArMs and to assess the impact of second sphere mutations on the catalytic tendencies. (Chapter 4) • Using the same approach, calculations were carried out for the in silico design of hydratases, but in this case based on the inclusion of a novel unnatural amino acid. This was first applied to the LmrR scaffold, for which mutants suggested via computation for optimum enantiomeric excess (ee) were then experimentally assessed with success. Next, based on the experience obtained, we expanded the de novo exercise towards the design of Artificial Metallopeptides. (Chapter 5) • The final part of the work focused on deciphering molecular variables that our previous studies showed to be far more complex than expected. This was the impact of the active site configuration to define the catalytic activity of the ArMs. In particular, we decoded the rearrangement of a variety of LmrR-heme complex for the cyclopropanation reaction to proceed. From this study it clearly appeared that the flexibility of the receptor is key for the porphyrin based ArMs to reach their catalytic activity. To further assess the importance of this molecular variable, we expanded this work to the study of distinct naturally occurring heme binding proteins. (Chapter 6) Overall, this Ph.D. represents a step forward on the methodological development of the computer based enzyme design. Furthermore, it sheds light on how transition metal compounds could cooperate with biological scaffolds at the molecular level with the focus on the de novo design of new biocatalysts.
