1
|
Rocha REO, Mariano DCB, Almeida TS, CorrêaCosta LS, Fischer PHC, Santos LH, Caffarena ER, da Silveira CH, Lamp LM, Fernandez-Quintero ML, Liedl KR, de Melo-Minardi RC, de Lima LHF. Thermostabilizing mechanisms of canonical single amino acid substitutions at a GH1 β-glucosidase probed by multiple MD and computational approaches. Proteins 2023; 91:218-236. [PMID: 36114781 DOI: 10.1002/prot.26424] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/26/2022] [Revised: 09/01/2022] [Accepted: 09/06/2022] [Indexed: 01/07/2023]
Abstract
β-glucosidases play a pivotal role in second-generation biofuel (2G-biofuel) production. For this application, thermostable enzymes are essential due to the denaturing conditions on the bioreactors. Random amino acid substitutions have originated new thermostable β-glucosidases, but without a clear understanding of their molecular mechanisms. Here, we probe by different molecular dynamics simulation approaches with distinct force fields and submitting the results to various computational analyses, the molecular bases of the thermostabilization of the Paenibacillus polymyxa GH1 β-glucosidase by two-point mutations E96K (TR1) and M416I (TR2). Equilibrium molecular dynamic simulations (eMD) at different temperatures, principal component analysis (PCA), virtual docking, metadynamics (MetaDy), accelerated molecular dynamics (aMD), Poisson-Boltzmann surface analysis, grid inhomogeneous solvation theory and colony method estimation of conformational entropy allow to converge to the idea that the stabilization carried by both substitutions depend on different contributions of three classic mechanisms: (i) electrostatic surface stabilization; (ii) efficient isolation of the hydrophobic core from the solvent, with energetic advantages at the solvation cap; (iii) higher distribution of the protein dynamics at the mobile active site loops than at the protein core, with functional and entropic advantages. Mechanisms i and ii predominate for TR1, while in TR2, mechanism iii is dominant. Loop A integrity and loops A, C, D, and E dynamics play critical roles in such mechanisms. Comparison of the dynamic and topological changes observed between the thermostable mutants and the wildtype protein with amino acid co-evolutive networks and thermostabilizing hotspots from the literature allow inferring that the mechanisms here recovered can be related to the thermostability obtained by different substitutions along the whole family GH1. We hope the results and insights discussed here can be helpful for future rational approaches to the engineering of optimized β-glucosidases for 2G-biofuel production for industry, biotechnology, and science.
Collapse
Affiliation(s)
- Rafael Eduardo Oliveira Rocha
- Laboratory of Molecular Modelling and Bioinformatics (LAMMB), Department of Physical and Biological Sciences, Campus Sete Lagoas, Universidade Federal de São João Del Rei, Sete Lagoas, Brazil.,Laboratory of Bioinformatics and Systems (LBS), Department of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil.,Laboratory of Molecular Modeling and Drug Design, Department of Biochemistry and Immunology, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
| | - Diego César Batista Mariano
- Laboratory of Bioinformatics and Systems (LBS), Department of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil
| | - Tiago Silva Almeida
- Laboratory of Molecular Modelling and Bioinformatics (LAMMB), Department of Physical and Biological Sciences, Campus Sete Lagoas, Universidade Federal de São João Del Rei, Sete Lagoas, Brazil
| | - Leon Sulfierry CorrêaCosta
- Laboratory of Molecular Modelling and Bioinformatics (LAMMB), Department of Physical and Biological Sciences, Campus Sete Lagoas, Universidade Federal de São João Del Rei, Sete Lagoas, Brazil.,Computational Modeling Coordination (COMOD), Laboratório Nacional de Computação Científica (LNCC), Petrópolis, Brazil
| | - Pedro Henrique Camargo Fischer
- Laboratory of Molecular Modelling and Bioinformatics (LAMMB), Department of Physical and Biological Sciences, Campus Sete Lagoas, Universidade Federal de São João Del Rei, Sete Lagoas, Brazil
| | - Lucianna Helene Santos
- Laboratory of Bioinformatics and Systems (LBS), Department of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil.,Laboratory of Molecular Modeling and Drug Design, Department of Biochemistry and Immunology, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
| | | | | | - Leonida M Lamp
- Institute of General, Inorganic and Theoretical Chemistry, and Center for Chemistry and Biomedicine Innsbruck (CCB), University of Innsbruck, Innsbruck, Austria
| | - Monica Lisa Fernandez-Quintero
- Institute of General, Inorganic and Theoretical Chemistry, and Center for Chemistry and Biomedicine Innsbruck (CCB), University of Innsbruck, Innsbruck, Austria
| | - Klaus Roman Liedl
- Institute of General, Inorganic and Theoretical Chemistry, and Center for Chemistry and Biomedicine Innsbruck (CCB), University of Innsbruck, Innsbruck, Austria
| | - Raquel Cardoso de Melo-Minardi
- Laboratory of Bioinformatics and Systems (LBS), Department of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil
| | - Leonardo Henrique França de Lima
- Laboratory of Molecular Modelling and Bioinformatics (LAMMB), Department of Physical and Biological Sciences, Campus Sete Lagoas, Universidade Federal de São João Del Rei, Sete Lagoas, Brazil.,Institute of General, Inorganic and Theoretical Chemistry, and Center for Chemistry and Biomedicine Innsbruck (CCB), University of Innsbruck, Innsbruck, Austria
| |
Collapse
|