Introduction
Interest in carbohydrates has rapidly increased in recent years and
reflects the essential functions that oligosaccharides and glycoconjugates play in living cells and organisms. Glycosyltransferases that
are involved in selective biosynthesis of glycan structures compose
a group of enzymes that transfer glycosyl residues from a nucleoside
phosphate to other molecules [14]. The outcome of the reaction,
catalyzed by these enzymes, is the creation of a new glycosidic linkage. However, the catalytic mechanism of glycosyltransferases is still
not fully understood [59].
Understanding the mechanism of glycosyltransferases at the
atomic level is of vital importance in biomedical research.
Experimental studies provide essential and necessary information
concerning the reaction mechanism, thermodynamics, and kinetics; however, experimental data are often inadequate for determining the reaction mechanism at the atomic level. Most importantly
Thomas Ltteke and Martin Frank (eds.), Glycoinformatics, Methods in Molecular Biology, vol. 1273,
DOI 10.1007/978-1-4939-2343-4_29, Springer Science+Business Media New York 2015
489
490
Igor Tvaroka
Theory
A description of the bond forming and bond breaking processes
during the catalytic reaction requires quantum mechanical methods. However, QM methods are restricted to systems of up to a
few hundred atoms and are unsuitable for large systems such as
enzymes. In fact, they were used only on small (cluster) models of
the enzymatic reaction. Cluster models provided valuable information of the mechanistic characteristics of a reaction catalyzed by
glycosyltransferases [1013]. However, a more realistic description
of reaction mechanism requires including the complete enzyme
environment in the computations. Thus, there is a need for the
method capable of treating up to several thousand atoms.
The field of mixed QM/MM methods has emerged during the
last few decades driven by needs of nanomaterials, condensed-phase
reactions, catalytic systems, and structural biology. A practical and
attractive solution to the energetic characterization of enzymatic
reaction is to combine both methods and utilize a QM treatment
for the active site (reaction) region and an MM method for the
remainder of the enzymatic environment. The QM/MM methods
combine the accuracy of a QM description with the low cost of
MM. In this chapter, we begin with a brief introduction to the basic
characteristics of the QM/MM approach. Then we introduce some
programs available for these calculations but do not review the
details. The following section describes a modeling process from
preparing structural and computational models to calculations of
reaction pathways. Finally, we illustrate this process showing the
key insights gained from applying the QM/MM method on the
catalytic mechanisms of the inverting glucosyltransferase.
2.1
QM/MM Method
In this section, we briefly describe only basic principles. The theoretical development and applications of QM/MM methods have
been recently discussed in a number of papers [1419] and readers
should refer to those reviews for more details. The general idea of
the QM/MM method is partitioning of the entire system into
localized regions, namely, QM and MM, as illustrated in Fig. 1.
For enzymatic reactions studies, the QM region is usually centered
on the active site where a chemical reaction proceeds. The QM
region includes structural moieties involved in the catalytic
reactions and amino acids involved in interactions with the reacting
491
(1)
492
Igor Tvaroka
493
3.2
QM/MM Partition
494
Igor Tvaroka
The first requirement for modeling a catalytic reaction of glycosyltransferases is the calculation of a potential energy surface (PES)
that accurately describes the chemical changes with the redistribution of electrons. In principle, the PES of the entire system (enzyme
and substrates) describes many TS and other stationary states
because there are many degrees of freedom that are not directly
related to the chemical changes but determine the shape of the
PES. Therefore, the progress of a catalytic reaction is often
described by the structural changes of a reduced number of reaction coordinates. The choice of reaction coordinate(s) is extremely
important. An improper choice of the reaction coordinate can bias
the calculation and yield slower convergence.
The catalytic reaction of glycosyltransferases can be characterized as the creation of one new glycosidic linkage between the
acceptor hydroxyl group oxygen and the anomeric carbon C1 of
the donor, cleavage of the donor glycosidic linkage between a
sugar and nucleotide diphosphate, and removal of a proton H
from the acceptor hydroxyl OH by a catalytic base. Thus, the reaction mechanism can be monitored by three reaction coordinates.
The first reaction coordinate r1 represents the nucleophilic attack
of the hydroxyl oxygen from the acceptor on the anomeric carbon
of the donor leading to the formation of a new -glycosidic linkage
495
496
Igor Tvaroka
497
The second is the interaction between HO6 from the acceptor and
-phosphate oxygen that stabilizes a negative charge on the
-pyrophosphate during the scission of the glycosidic linkage.
The calculated TS structure displays features (Fig. 4) characteristic of SN2 reaction. The nearly simultaneous nucleophilic addition
and dissociation steps are demonstrated by a distance of 1.912
and 2.542 for the forming and breaking glycosidic bond, respectively. The glucopyranose ring of GlcNAc undergoes significant
structural changes progressing from the ES to TS model. The
GlcNAc ring adopts the 4H3 conformation in the TS model with
almost planar arrangement around anomeric carbon, whereas in the
ES, the ring conformation is the 4C1 chair. This is accompanied by
a shortening of the ring oxygenanomeric carbon (C1O5) bond
length to 1.333 . This partial double-bond character of the C1
O5 bond is a result of delocalization of the O5 lone pairs into C1
and contributes to sp2 hybridization at the anomeric carbon. The
lone pairs delocalization also stabilizes the partial positive charge
on the anomeric carbon C1 and facilitates simultaneous interactions of C1 with the leaving group and attacking nucleophile. It is
noteworthy that cleavage of the C1O1 glycosidic bond is accompanied by the 17 rotation of the -phosphate oxygen. The calculated TS structure was supported by a normal-mode analysis of the
QM part of the model and, together with the calculated -deuterium
kinetics effect (KIE), supported an SN2-type mechanism.
498
Igor Tvaroka
Acknowledgements
This work was supported by the Scientific Grant Agency of the
Ministry of Education of Slovak Republic and Slovak Academy of
Sciences (grants VEGA-02/0176/09 and VEGA-02/0101/11).
References
1. Beyer TA, Sadler JE, Rearick JI, Paulson JC,
Hill RL (1981) Glycosyltransferases and their
use in assessing oligosaccharide structure and
structure-function relationships. Adv Enzymol
52:23175
2. Schachter H (1991) Enzymes associated with
glycosylation. Curr Opin Struct Biol 1:
755765
3. Kleene R, Berger EG (1993) The molecular
and cell biology of glycosyltransferases.
Biochim Biophys Acta 1154:283325
4. Montreuil J, Vliegenthart JFG, Schachter H
(1995) Glycoproteins. In: Neuberger A, van
Deenen LLM (eds) New comprehensive biochemistry, vol 29a. Elsevier, Amsterdam
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
499
http://www.schrodinger.com
http://www.amber.scripps.edu
http://www.charmm.org
http://www.cse.clrc.ac.uk/qcg/chemshell
http://www.pdb.org
Modeller, A program for protein structure
modeling, http://salilab.org/modeller/modeller.html
Pettersen EF, Goddard TD, Huang CC,
Couch GS, Greenblatt DM, Meng EC, Ferrin
TE (2004) J Comput Chem 25:1605
(2009) Maestro version 9.0, New York, NY:
Schrdinger, LLC
Schrdinger Suite (2009) Protein preparation
wizard, Epik version 2.0, New York, NY:
Schrdinger, LLC
Morris GM, Lim-Wilby M (2008) Molecular
docking. In: Kukol A (ed) Methods in molecular biology, vol 443. Humana Press, Totowa,
NJ, pp 365382
Sumovski CV, Ochsenfeld C (2009) A convergence study of QM/MM isomerization energies with the selected size of the QM |region
for peptidic systems. J Phys Chem A
113:1173411741
Hu LH, Sderhjelm P, Ryde U (2011) On the
convergence of QM/MM energies. J Chem
Theor Comput 7:761777
Xu X, Alecu IM, Truhlar DG (2011) How well
can modern density functionals predict internuclear distances at transition state? J Chem
Theor Comput 7:16671676
Ka J, Tvaroka I (2009) Comparative DFT
study on the glycosidic bond in reactive species
of galactosyl diphosphates. Chem Papers
63:598607
Kozmon S, Tvaroka I (2006) Catalytic mechanism of glycosyltransferases: hybrid quantum
mechanical/molecular mechanical study of the
Inverting N-acetylglucosaminyltransferase I.
J Am Chem Soc 128:1692116927
Unligil UM, Zhou S, Yuwaraj S, Sarkar M,
Schachter H, Rini JM (2000) X-ray crystal
structure of rabbit N-acetylglucosaminyltransferase I: catalytic mechanism and a new
protein superfamily. EMBO J 19:52695280
Schachter H, Reck F, Paulsen H (2003) Use of
synthetic oligosaccharide substrate analogs to
map the active sites of N-acetylglucosaminyltransferases I and II. Methods Enzymol 363:
459475
(2009) Glide version 5.5, New York, NY:
Schrdinger, LLC
SCM (2005) ADF 2005.01, Amsterdam
Cornell WD, Cieplak P, Bayly CI, Gould IR,
Merz KM, Ferguson DM, Spellmeyer DC, Fox T,
Caldwell JW, Kollman PA (1995) J Am Chem
Soc 117(19):51795197