522
T. Borowski and M. Szaleniec
In order to further illustrate the degree to which the oxidized Moco is energized
by the twist introduced by the enzyme, one can calculate relaxation energy of the
ES upon removal of constrains. For EBDH, such difference was estimated to be
14.3 kcal/mol [81], while for S25DH it was 14.9 kcal/mol for the smaller model
used by Rugor et al. and 16.8 kcal/mol and 17.7 kcal/mol for extended model 1 and
model 2, respectively. Similarly, it has been calculated that in the case of arsenate
oxidase, the energy difference between equilibrium and strained geometry is in the
range of 12.9 kcal/mol [80].
The above-discussed results clearly illustrate that in certain cases, where the cofactor is fairly rigid and extended, the steric effects imposed by the protein can play
non-negligible role for the catalytic process. Hence, the origin of the cluster model,
i.e. if it comes from X-ray structure for oxidized or reduced form of the enzyme,
its size and the way constraints are introduced into it all will have influence on the
energetic profile for the studied reaction.
10 Conclusions
Computational studies on metalloenzymes can provide valuable new insights into
structure and reaction mechanisms, yet their credibility depends on multiple factors.
In this contribution, we have tried to illustrate the key decisions and steps one has to
take in such projects (validating X-ray structures, re-refining metal sites, determining
residues protonation state, modelling missing loop regions or whole proteins by comparative modelling, performing MD simulations, choosing representative snapshots
and constructing QM models). While this list is by no means exhaustive, we hope
that when the content of this chapter will be combined with the material covered in
other chapters of this book, it will offer a more comprehensive view on modelling
metalloenzymes.
Acknowledgements This research was supported in part by PL-Grid Infrastructure. Computations
were performed in the AGH Cyfronet Supercomputer Centre. Authors acknowledge financial support of the statutory research fund of ICSC PAS. M. Sz. acknowledges the input of Ewelina Biała,
the undergraduate student conducting modeling of S25DH under supervision of M. Sz.
References
1. Nitschke W, McGlynn SE, Milner-White EJ, Russell MJ (2013) Biochim Biophys Acta Bioenerg 1827:871
2. Lovell T (2003) Coord Chem Rev 238–239:211
3. Andreini C, Bertini I, Rosato A (2009) Acc Chem Res 42:1471
4. Andreini C, Bertini I, Cavallaro G, Holliday GL, Thornton JM (2008) JBIC J Biol Inorg Chem
13:1205
5. Jernigan R, Raghunathan G, Bahar I (1994) Curr Opin Struct Biol 4:256
T. Borowski and M. Szaleniec
In order to further illustrate the degree to which the oxidized Moco is energized
by the twist introduced by the enzyme, one can calculate relaxation energy of the
ES upon removal of constrains. For EBDH, such difference was estimated to be
14.3 kcal/mol [81], while for S25DH it was 14.9 kcal/mol for the smaller model
used by Rugor et al. and 16.8 kcal/mol and 17.7 kcal/mol for extended model 1 and
model 2, respectively. Similarly, it has been calculated that in the case of arsenate
oxidase, the energy difference between equilibrium and strained geometry is in the
range of 12.9 kcal/mol [80].
The above-discussed results clearly illustrate that in certain cases, where the cofactor is fairly rigid and extended, the steric effects imposed by the protein can play
non-negligible role for the catalytic process. Hence, the origin of the cluster model,
i.e. if it comes from X-ray structure for oxidized or reduced form of the enzyme,
its size and the way constraints are introduced into it all will have influence on the
energetic profile for the studied reaction.
10 Conclusions
Computational studies on metalloenzymes can provide valuable new insights into
structure and reaction mechanisms, yet their credibility depends on multiple factors.
In this contribution, we have tried to illustrate the key decisions and steps one has to
take in such projects (validating X-ray structures, re-refining metal sites, determining
residues protonation state, modelling missing loop regions or whole proteins by comparative modelling, performing MD simulations, choosing representative snapshots
and constructing QM models). While this list is by no means exhaustive, we hope
that when the content of this chapter will be combined with the material covered in
other chapters of this book, it will offer a more comprehensive view on modelling
metalloenzymes.
Acknowledgements This research was supported in part by PL-Grid Infrastructure. Computations
were performed in the AGH Cyfronet Supercomputer Centre. Authors acknowledge financial support of the statutory research fund of ICSC PAS. M. Sz. acknowledges the input of Ewelina Biała,
the undergraduate student conducting modeling of S25DH under supervision of M. Sz.
References
1. Nitschke W, McGlynn SE, Milner-White EJ, Russell MJ (2013) Biochim Biophys Acta Bioenerg 1827:871
2. Lovell T (2003) Coord Chem Rev 238–239:211
3. Andreini C, Bertini I, Rosato A (2009) Acc Chem Res 42:1471
4. Andreini C, Bertini I, Cavallaro G, Holliday GL, Thornton JM (2008) JBIC J Biol Inorg Chem
13:1205
5. Jernigan R, Raghunathan G, Bahar I (1994) Curr Opin Struct Biol 4:256