Intrinsic Bending of Microtubule Protofilaments
Andrea Grafmu ¨ller1and Gregory A. Voth1,*
1Department of Chemistry, James Franck Institute, and Computation Institute, University of Chicago, 5735 S. Ellis Avenue, Chicago,
IL 60637, USA
The complex polymerization dynamics of the micro-
tubule (MT) plus end are closely linked to the hydro-
lysis of the GTP nucleotide bound to the b-tubulin.
The destabilization is thought to be associated with
the conformational change of the tubulin dimers
from the straight conformation in the MT lattice to
a curved conformation. It remains under debate
whether this transformation is directly related to the
nucleotide state, or a consequence of the longitu-
dinal or lateral contacts in the MT lattice. Here, we
present large-scale atomistic simulations of short
tubulin protofilaments with both nucleotide states,
starting from both extreme conformations. Our simu-
lations indicate that both interdimer and intradimer
contacts in both GDP and GTP-bound tubulin dimers
and protofilaments in solution bend. There are no
observable differences between the mesoscopic
properties of the contacts in GTP and GDP-bound
tubulin or the intradime and interdimer interfaces.
Microtubules (MTs) are rigid cylindrical filaments assembled
from ab-tubulin heterodimers. These heteropolymers form one
of the key components of the cytoskeleton, and are involved in
trafficking, structural support, and cytokinesis (Hyams and
Lloyd, 1993). Their functionality is closely linked to the complex
polymerization dynamics at the MT plus end (with the b-tubulin
exposed) that switch between phases of growth and rapid
disassembly (Mitchison and Kirschner, 1984). This so-called
‘‘dynamic instability’’ is controlled by the hydrolysis of GTP in
b-tubulin upon polymerization (Howard and Hyman, 2003).
Although vitally is important, the underlying molecular mecha-
nism leading to the dynamic instability is not well understood.
GTP hydrolysis is required for dynamic MTs, and caps of GTP-
bound tubulin stabilizing the MT plus end have been observed
experimentally both in vitro (Drechsel and Kirschner, 1994;
Caplow and Shanks, 1996; Desai and Mitchison, 1997) and
in vivo (Dimitrov et al., 2008). When these caps are hydrolyzed,
rapid depolymerization takes place.
Explanations for the role of GTP hydrolysis in the polymeriza-
tion dynamics are based on the two available crystal structures
for the tubulin dimer. One of these structures is a straight confor-
mation obtained from electron crystallography of taxol-stabi-
lized, zinc-induced tubulin sheets (Nogales et al., 1998; Lo ¨we
et al., 2001); the other is in a curved conformation bound to
a fragment of the stathmin homolog RB3 and colchicine (Gigant
et al., 2000; Ravelli et al., 2004). It is assumed that the straight
conformation is very similar to that in the MT lattice, and an esti-
mate for the conformation and contacts in the MT was obtained
by docking the atomistic structure of the straight conformation
into a lower resolution (8 A˚) image (Li et al., 2002).
These structures have led to two opposing models for
a possible mechanism by which GTP promotes polymerization
and GTP hydrolysis leads to disassembly. The allosteric model
(Wang and Nogales, 2005; Nogales and Wang, 2006) postulates
that GTP-bound ab-tubulin has a substantially straighter confor-
mation than the curved GDP-tubulin and would, therefore, be
(Buey et al., 2006; Rice et al., 2008), on the other hand, predicts
that both conformations are bent, and the conformational
change is a consequence of integration into the lattice structure,
rather than the cause. The role of the nucleotide in this model is
to alter the strength of the lateral contacts. The essential differ-
ence between these two models is the unconstrained conforma-
tion of GTP-bound tubulin dimers and protofilaments in solution.
Whereas the allosteric model predicts a straight, or straighter,
conformation than that of GDP-bound tubulin, the lattice model
postulates a similar curvature for both nucleotide states.
Experimental evidence exists for both models but is indirect,
so that the question remains unresolved. MT ends grow in
sheet-like assemblies but disassemble in ringlike structures
(Chre ´tien et al., 1995; Arnal et al., 2000) that can be stabilized
by a number of agents (Nogales et al., 2003; Wang and Nogales,
2005; Elie-Caille et al., 2007). In these structures the curvature of
tubulin bound to GMPCPP, a nonhydrolyzable GTP analog, was
agreement with the allosteric model. Support for the lattice
model stems from crystal structures of g-tubulin (Aldaz et al.,
2005; Rice et al., 2008) as well as several more distantly related
bacterial tubulin homologs (Gigant et al., 2005; Schlieper et al.,
this model (Manuel Andreu et al., 1989; Shearwin and Timasheff,
1994; Rice et al., 2008).
Because this evidence is either indirect or involves external
influences such as destabilizing agents or large protein assem-
blies imposing geometric constraints that may affect the intrinsic
bending of the dimer, both models remain under debate.
Recent experiments on the size distribution of tubulin oligo-
mers suggest that the majority of tubulin dimers in solution exist
as short protofilaments with no discernable difference between
GTP-bound and GDP-bound protofilaments (Mozziconacci
Structure 19, 409–417, March 9, 2011 ª2011 Elsevier Ltd All rights reserved 409
et al., 2008). The GTP analog GMPCPP, on the other hand, did
change the longitudinal contacts, suggesting that GMPCPP-
bound tubulin may not be an optimal model for GTP-bound
Computational studies may help to shed light on the question
external factors. Instead, simulations are limited by the small
length and timescales they can sample and have to rely on the
accuracy of the underlying force field. Thus, the large size and
long timescale dynamics of tubulin aggregates present a
challenge for atomistic simulations. Various simplified models
have helped to understand properties of tubulin and MTs
(Deriu et al., 2010; Baker et al., 2001; Keskin et al., 2002;
VanBuren et al., 2002, 2005; Sept et al., 2003; Tuszy? nski et al.,
2005; Drabik et al., 2007; Dima and Joshi, 2008; Jiang et al.,
2008; Neek-Amal et al., 2008), and identify binding sites (Mitra
and Sept, 2004, 2006; Morrissette et al., 2004) and conforma-
tions of the termini (Luchko et al., 2008). All-atom molecular
dynamics (MD) simulations of single tubulin dimers have sug-
gested an intrinsic bending of the dimers for both nucleotides,
with different angle and direction for GTP and GDP (Gebremi-
chael et al., 2008). A second simulation study observed bending
in a different direction and further concluded that dimer flexibility
may play a more important role than the actual bending angle
(Bennett et al., 2009). Further MD simulations using small pieces
of the MT lattice have investigated the effects of taxol binding
(Mitra and Sept, 2008) and the elastic properties of tubulin in
the MT lattice (Sept and MacKintosh, 2010; Wells and Aksimen-
Here, we present atomistic MD simulations of short protofila-
ments in atomistic resolution. Such aggregates are likely to
conformations of intradimer and interdimer contacts. Simula-
tions are started from both the straight structure, which is often
GDP and GTP-bound tubulin filaments in solution have a very
similar curved conformation.
RESULTS AND DISCUSSION
Large-scale properties that are comparable to experimentally
accessible properties can be analyzed by treating the filaments
as stiff polymers, where each monomer is represented by
a bead at its center of mass, as illustrated in Figure 1A.
Observing the trajectories at this scale, we find that after
20–50 ns the straight protofilaments start to assume clearly
curved conformations. Several of the final snapshots shown in
Figure 1B are as strongly bent as the curved RB3-SLD structure,
albeit in a somewhat different direction from that seen in the
structure. The snapshots of the curved protofilaments shown
in Figure 1C have similar curvatures. However, several bend in
different directions and have a ‘‘twisted’’ appearance. In some
trajectories the curvature is almost in the opposite direction.
A more detailed look at the atomistic structure shows that this
heterogeneity is due to the interdimer contacts in the stathmin-
bound structure, which have a less-stable conformation, as will
be discussed later.
Inspection of the distribution of the radius of curvature seen in
the last 20 ns of the simulations, shown in Figure 1D, reveals that
simulations starting from the curved conformation for both
nucleotides as well as the GTP-bound filaments starting from
Figure 1. Tubulin Conformations
(A) Molecular model together with the cg representation where each monomer
is represented by a bead at its center of mass.
(BandC) Snapshotsofthe MDtrajectories. GDP-boundtubulin (red) and GTP-
bound tubulin (green) protofilaments are shown together with the 1JFF (gray)
and RB3 (black) crystal structures.
(B) Trajectories starting from the straight conformation.
(C) Trajectories starting from the curved structure.
(D) The distribution of curvatures from the last 20 ns of the trajectories.
(E) Superimposed b-monomers from the straight (light blue) and curved (dark
blue) crystal structures, together with a simulation snapshot of a trajectory
starting from the straight structure (yellow, red) seen from the top (left) and
side (right). The structures were superimposed by fitting the terminal domains.
In the simulation snapshots the intermediatedomain (red) has shifted, to agree
well with the curved structure.
Intrinsic Bending of Microtubule Protofilaments
410 Structure 19, 409–417, March 9, 2011 ª2011 Elsevier Ltd All rights reserved
Manuel Andreu, J., Garcia de Ancos, J., Starling, D., Hodgkinson, J.L., and
Bordas, J. (1989). A synchrotron X-ray scattering characterization of purified
tubulinand ofits expansion induced
Biochemistry 28, 4036–4040.
Martyna, G.J., Tobias, D.J., and Klein, M.L. (1994). Constant pressure molec-
ular dynamics algorithms. J. Chem. Phys. 101, 4177–4189.
Mickey, B., and Howard, J. (1995). Rigidity of microtubules is increased by
stabilizing agents. J. Cell Biol. 130, 909–917.
Mitchison, T., and Kirschner, M. (1984). Dynamic instability of microtubule
growth. Nature 312, 237–242.
Mitra, A., and Sept, D. (2004). Localization of the antimitotic peptide and
depsipeptide binding site on beta-tubulin. Biochemistry 43, 13955–13962.
Mitra, A., and Sept, D. (2006). Binding and interaction of dinitroanilines with
apicomplexan and kinetoplastid alpha-tubulin. J. Med. Chem. 49, 5226–5231.
Mitra, A., and Sept, D. (2008). Taxol allosterically alters the dynamics of the
tubulin dimer and increases the flexibility of microtubules. Biophys. J. 95,
Morrissette, N.S., Mitra, A., Sept, D., and Sibley, L.D. (2004). Dinitroanilines
bind alpha-tubulin to disrupt microtubules. Mol. Biol. Cell 15, 1960–1968.
Mozziconacci, J., Sandblad, L., Wachsmuth, M., Brunner, D., and Karsenti, E.
(2008). Tubulin dimers oligomerize before their incorporation into microtu-
bules. PLoS ONE 3.
Needleman, D.J., Ojeda-Lopez, M.A., Raviv, U., Ewert, K., Miller, H.P., Wilson,
L., and Safinya, C.R. (2005). Radial compression of microtubules and
the mechanism of action of taxol and associated proteins. Biophys. J. 89,
Neek-Amal, M., Radja, N.H., and Ejtehadi, M.R. (2008). Effective potential of
longitudinal interactions between microtubule protofilaments. Phys. Rev.
E Stat. Nonlin. Soft Matter Phys. 78.
Newstead, S., Fowler, P.W., Bilton, P., Carpenter, E.P., Sadler, P.J.,
Campopiano, D.J., Sansom, M.S.P., and Iwata, S. (2009). Insights into how
nucleotide-binding domains power ABC transport. Structure 17, 1213–1222.
Nogales, E., and Wang, H.W. (2006). Structural mechanisms underlying
nucleotide-dependent self-assembly of tubulin and its relatives. Curr. Opin.
Struct. Biol. 16, 221–229.
Nogales, E., Wolf, S.G., and Downing, K.H. (1998). Structure of the alpha beta
tubulin dimer by electron crystallography. Nature 391, 199–203.
Nogales, E., Wang, H.W., and Niederstrasser, H. (2003). Tubulin rings: which
way do they curve? Curr. Opin. Struct. Biol. 13, 256–261.
Oliva, M.A., Trambaiolo, D., and Lo ¨we, J. (2007). Structural insights into the
conformational variability of FtsZ. J. Mol. Biol. 373, 1229–1242.
Phillips, J.C., et al. (2005). Scalable molecular dynamics with NAMD.
J. Comput. Chem. 26, 1781–1802.
Ravelli, R.B.G., Gigant, B.,Curmi, P.A., Jourdain, I., Lachkar, S., Sobel, A.,and
Knossow, M. (2004). Insight into tubulin regulation from a complex with colchi-
cine and a stathmin-like domain. Nature 428, 198–202.
Rice, L.M., Montabana, E.A., and Agard, D.A. (2008). The lattice as allosteric
effector: Structural studies of alpha beta- and gamma-tubulin clarify the role
of GTP in microtubule assembly. Proc. Natl. Acad. Sci. USA 105, 5378–5383.
Ryckaert, J., Ciccotti, G., and Berendsen, H. (1977). Numerical-integration of
Cartesian equations of motion of a system with constraints: molecular
dynamics of N-alkanes. J. Comput. Phys. 23, 327–341.
Schek, H.T., Gardner, M.K., Cheng, J., Odde, D.J., and Hunt, A.J. (2007).
Microtubule assembly dynamics at the nanoscale. Curr. Biol. 17, 1445–1455.
Schlieper, D., Oliva, M.A., Andreu, J.M., and Lo ¨we, J. (2005). Structure of
bacterial tubulin BtubA/B: Evidence for horizontal gene transfer. Proc. Natl.
Acad. Sci. USA 102, 9170–9175.
Sept, D., and MacKintosh, F.C. (2010). Microtubule elasticity: connecting all-
atom simulations with continuum mechanics. Phys. Rev. Lett. 104, 018101.
Sept, D., Baker, N.A., and McCammon, J.A. (2003). The physical basis of
microtubule structure and stability. Protein Sci. 12, 2257–2261.
Shearwin, K.E., and Timasheff, S.N. (1994). Effect of colchicine analogs on the
dissociation of alpha-beta-tubulin into subunits: the locus of colchicine
binding. Biochemistry 33, 894–901.
Tuszy? nski, J.A., Brown, J.A., Crawford, E., Carpenter, E.J., Nip, M.L.A., Dixon,
J.M., and Satari? c, M.V. (2005). Molecular dynamics simulations of tubulin
structure and calculations of electrostatic properties of microtubules. Math.
Comput. Model. 41, 1055–1070.
VanBuren, V., Odde, D.J., and Cassimeris, L. (2002). Estimates of lateral and
longitudinal bond energies within the microtubule lattice. Proc. Natl. Acad.
Sci. USA 99, 6035–6040.
VanBuren, V., Cassimeris, L., and Odde, D.J. (2005). Mechanochemical model
of microtubule structure and self-assembly kinetics. Biophys. J. 89, 2911–
Venier, P., Maggs, A.C., Carlier, M.F., and Pantaloni, D. (1994). Analysis of
microtubule rigidity using hydrodynamic flow and thermal fluctuations.
J. Biol. Chem. 269, 13353–13360.
Wang, H.W., and Nogales, E. (2005). Nucleotide-dependent bending flexibility
of tubulin regulates microtubule assembly. Nature 435, 911–915.
Wassenaar, T.A., and Mark, A.E. (2006). The effect of box shape on the
dynamic properties of proteins simulated under periodic boundary conditions.
J. Comput. Chem. 27, 316–325.
Wells, D.B., and Aksimentiev, A. (2010). Mechanical properties of a complete
microtubule revealed through molecular dynamics simulation. Biophys. J. 99,
Intrinsic Bending of Microtubule Protofilaments
Structure 19, 409–417, March 9, 2011 ª2011 Elsevier Ltd All rights reserved 417