Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Computational Studies of Nitrogenase

Cao, Lili LU (2020)
Abstract
Abstract
Nitrogenase is the only enzyme that can convert the inert nitrogen molecule to ammonia, so that it can be used for biomass production and in biosynthetic pathways. It contains a complicated acitve site, composed of eight metal ions, nine sulfur ions and one carbide ion (the FeMo cluster). Although it has been thoroughly studied with crystallographic, biochemical, kinetic, spectroscopic and computational methods, the reaction mechanism is still not known and many conflicting hypotheses have been presented. To solve some of these problems, we have performed a thorough and systematic study of nitrogenase with various computational approaches, based on a combination of quantum mechanics (QM), molecular mechanics (MM) and sometimes... (More)
Abstract
Nitrogenase is the only enzyme that can convert the inert nitrogen molecule to ammonia, so that it can be used for biomass production and in biosynthetic pathways. It contains a complicated acitve site, composed of eight metal ions, nine sulfur ions and one carbide ion (the FeMo cluster). Although it has been thoroughly studied with crystallographic, biochemical, kinetic, spectroscopic and computational methods, the reaction mechanism is still not known and many conflicting hypotheses have been presented. To solve some of these problems, we have performed a thorough and systematic study of nitrogenase with various computational approaches, based on a combination of quantum mechanics (QM), molecular mechanics (MM) and sometimes also crystallographic refinment. We have:
• Decided the protonation states of eight key amino acid residues around the active site and showed that the homocitrate ligand is singly prontated on the hydroxide group.
• Studied how the the broken-symmetry (BS) state for the FeMo cluster depends on the QM method, the basis sets, the surrounding protein and the protonation and oxidation state of the cluster. Thereby, we could propse a practical prodecure to deal with these states in computational studies.
• Predicted the most stable protonation state for the E0–E4 states of nitrogenase with two density functional theory (DFT) methods: The TPSS functional perfers the protonation of Fe, whereas B3LYP prefers protonation of the carbide ion.
• Showed that different DFT methods give relative eneriges that can differ by ~1100 kJ/mol for nitrogenase. This is the main reason for the diverging computational results. Pure functionals and TPSSh predict the best geometries of the E0 state, whereas B3LYP and PBE0 give more reliable H2 dissociation energies.
• Showed that the most stable E4 structure obtained with pure functionals has two hydride ions bridging between two pairs of iron ions and two protons on the sulfide ions, in agreement with experiments. We also found a new low-energy BS state in which only two Fe ions have minority spin.
• Predicted that the most stable binding mode of N2H2 to nitrogenase involves trans-HNNH binding to Fe2. However, other binding modes, e.g. involving cis-HNNH or NNH2 and binding to Fe6, are rather low in energy.
• Suggested an alternating reaction mechanism for nitrogenase with a dissociated S2B ligand.
• Decided the most stable BS states for the P-cluster in four oxidation states and showed that the one-electron oxidised state involves a protonated Cys-88, but a deprotonated Ser-188.
• Developed a novel quantum-refinement approach allowing for disorder in the QM system and applied it to the P-cluster in two crystal structures of nitrogenase.
• Shown by quantum refinement that a recent crystal structure of V-nitrogenase does not involve a N-derived ligand, but rather a hydride-inhibited state with a significant amount of the undissociated S2B ligand.
Together, these studies have taken us a significant step closer to an atomic understanding of nitrogenase, showing that experiments and calculations start to converge, illustrating the strength of quantum refinement, but also pointing out important problems that need to be solved in computational studies of nitrogenase.
(Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Dr. Björnsson, Ragnar, MPI for Chemical Energy Conversion, Mülheim an der Ruhr, Germany
organization
publishing date
type
Thesis
publication status
published
subject
keywords
nitrogenase, QM/MM, quantum refinement, FeMo cluster, E4 state, N2 binding, V-nitrogenase, P-cluster, S2B dissociation, broken-symmetry state, homocitrate, DFT, protonation state, disorder, ComQumX-2QM
pages
290 pages
publisher
Lund University (Media-Tryck)
defense location
Lecture Hall A, Kemicentrum, Naturvetarvägen 14, Lund (Live streaming: https://youtu.be/exZaUM4VTeg)
defense date
2020-05-29 13:00:00
ISBN
978-91-7422-741-3
978-91-7422-742-0
language
English
LU publication?
yes
id
977929dc-cf88-4e87-9f37-4151eb8483b1
date added to LUP
2020-05-05 15:09:04
date last changed
2020-05-08 19:14:45
@phdthesis{977929dc-cf88-4e87-9f37-4151eb8483b1,
  abstract     = {{Abstract<br/>Nitrogenase is the only enzyme that can convert the inert nitrogen molecule to ammonia, so that it can be used for biomass production and in biosynthetic pathways. It contains a complicated acitve site, composed of eight metal ions, nine sulfur ions and one carbide ion (the FeMo cluster). Although it has been thoroughly studied with crystallographic, biochemical, kinetic, spectroscopic and computational methods, the reaction mechanism is still not known and many conflicting hypotheses have been presented. To solve some of these problems, we have performed a thorough and systematic study of nitrogenase with various computational approaches, based on a combination of quantum mechanics (QM), molecular mechanics (MM) and sometimes also crystallographic refinment. We have: <br/>•	Decided the protonation states of eight key amino acid residues around the active site and showed that the homocitrate ligand is singly prontated on the hydroxide group. <br/>•	Studied how the the broken-symmetry (BS) state for the FeMo cluster depends on the QM method, the basis sets, the surrounding protein and the protonation and oxidation state of the cluster. Thereby, we could propse a practical prodecure to deal with these states in computational studies.<br/>•	Predicted the most stable protonation state for the E0–E4 states of nitrogenase with two density functional theory (DFT) methods: The TPSS functional perfers the protonation of Fe, whereas B3LYP prefers protonation of the carbide ion.<br/>•	Showed that different DFT methods give relative eneriges that can differ by ~1100 kJ/mol for nitrogenase. This is the main reason for the diverging computational results. Pure functionals and TPSSh predict the best geometries of the E0 state, whereas B3LYP and PBE0 give more reliable H2 dissociation energies.<br/>•	Showed that the most stable E4 structure obtained with pure functionals has two hydride ions bridging between two pairs of iron ions and two protons on the sulfide ions, in agreement with experiments. We also found a new low-energy BS state in which only two Fe ions have minority spin.<br/>•	Predicted that the most stable binding mode of N2H2 to nitrogenase involves trans-HNNH binding to Fe2. However, other binding modes, e.g. involving cis-HNNH or NNH2 and binding to Fe6, are rather low in energy. <br/>•	Suggested an alternating reaction mechanism for nitrogenase with a dissociated S2B ligand.<br/>•	Decided the most stable BS states for the P-cluster in four oxidation states and showed that the one-electron oxidised state involves a protonated Cys-88, but a deprotonated Ser-188.<br/>•	Developed a novel quantum-refinement approach allowing for disorder in the QM system and applied it to the P-cluster in two crystal structures of nitrogenase.<br/>•	Shown by quantum refinement that a recent crystal structure of V-nitrogenase does not involve a N-derived ligand, but rather a hydride-inhibited state with a significant amount of the undissociated S2B ligand.<br/>Together, these studies have taken us a significant step closer to an atomic understanding of nitrogenase, showing that experiments and calculations start to converge, illustrating the strength of quantum refinement, but also pointing out important problems that need to be solved in computational studies of nitrogenase.<br/>}},
  author       = {{Cao, Lili}},
  isbn         = {{978-91-7422-741-3}},
  keywords     = {{nitrogenase, QM/MM, quantum refinement, FeMo cluster, E4 state, N2 binding, V-nitrogenase, P-cluster, S2B dissociation, broken-symmetry state, homocitrate, DFT, protonation state, disorder, ComQumX-2QM}},
  language     = {{eng}},
  month        = {{04}},
  publisher    = {{Lund University (Media-Tryck)}},
  school       = {{Lund University}},
  title        = {{Computational Studies of Nitrogenase}},
  url          = {{https://lup.lub.lu.se/search/files/79350093/kappa.pdf}},
  year         = {{2020}},
}