Back to search

University of Strathclyde

Computational modelling of enzyme activity to speed up biocatalyst redesign

Abstract

dc:description.abstract

Biocatalysis is increasingly used for the synthesis of pharmaceuticals intermediates. However, to expand the applicability of these methods current timelines for biocatalyst optimization need to be reduced. Quantum Mechanics/Molecular Mechanics (QM/MM) methods allow, in principle, the accurate evaluation of enzymatic activities and thus offer an interesting option for the in silico pre-screening of variants. However, standard QM/MM methods are a computationally expensive class of methods and thus for practical large scale applications, approximations need to be made. In this work, a QM/MM-based protocol for hotspot identification has been developed and tested.The establishment and validation of an internal protocol for accurate QM/MM calculations was achieved through the mechanistic study of aldose reductase. This study highlights the importance of parameters, such as size of the QM region or the choice of the QM method, on differentiating between competitive mechanisms and consequently on accurately determining the role of the environment on the energy profile of the reaction.Having a validated QM/MM methodology, an enzymatic amide bond formation was used as a case study to elaborate and test a protocol for hotspot identification. In the resulting protocol three major approximations were introduced to speed up calculations: starting from a single snapshot, neglecting possible reorganizations consecutive to the mutation and focusing mainly on electrostatic effects. Two different types of charge modification protocols were investigated: charge deletion and charge introduction. From this study one specific hotspot was identified.In a further study, a homology model strategy was conducted to cope with the absence of experimentally determined structure, a frequent issue in enzyme design. Our previously established protocol was re-tested starting from the homology model and hotspots were identified. Finally, an evaluation of solvent free calculations, as an option to further accelerate the calculations, was also carried out. Encouraging results were obtained in the solvent free studies as similar hotspots were obtained relative to the water or toluene solvated models. Nonetheless, significant variations do exist between different solvents and further studies are necessary to validate the use of this approximation in a wider context.

Degree

thesis:*
Name dc:type.qualificationname
phd
Level dc:type.qualificationlevel
doctoral-pg
Grantor dc:publisher.institution
University of Strathclyde
Year dc:date.issued
2018

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Dréanic, Marie-Pierre
Advisors dc:contributor.advisor
  • Edge, Colin
  • Tuttle, Tell

Rights

Identifiers

dc:identifier.*
Identifier
T15955

Chain of custody

source
Harvested from
University of Strathclyde
Base URL
stax.strath.ac.uk/catalog/oai
Last updated
2026-11-20
Source record
OAI-PMH GetRecord
related terms
citation

Dréanic, Marie-Pierre. Computational modelling of enzyme activity to speed up biocatalyst redesign. doctoral-pg thesis, University of Strathclyde, 2018. https://stax.strath.ac.uk/concern/theses/4j03d015b