ABSTRACT
To study energetics and structural transitions at the atomistic
level in complex biological molecules such as DNA and proteins
computational modeling represents a powerful and widely used
approach [1]. Traditional experimental techniques, such as X-ray
crystallography and solution NMR, yield insights into these phe-
nomena but are plagued by a number of problems associated with
crystallization and resolution issues, as well as accessibility to short
lived high energy states and time domain information [2]. During
the past two decades, several all-atom empirical force fields (FF) for
biologicalmolecules have been developed, including CHARMM [3, 4],
AMBER [5], GROMOS [6], and OPLS [7], among others. They proved
to be remarkably useful for a range of systems that contain 10,000
or more atoms, being computationally cost-effective due to the
utilization of simplified potential energy functions for determination
of the energies and forces acting on such systems. One of the
major limitations of most of the current force fields associated
with such simplifications is the treatment of electrostatics within
the framework of the fixed-atomic-charge approximation, where
effective charges assigned to particles are independent of a system’s
configuration and are adjusted to account for the influence of in-
duced polarization in an average way. Such force fields are currently
used for most of biomolecular simulations and commonly termed
“additive” indicating they do not account for many-body induced
polarization effects explicitly. However, for many complex biological
systems such as polyanionic DNAor protein immersed in an aqueous
salt environment, whose conformational behavior is determined to
a significant extent by solvation effects and interactions with the
surrounding ionic atmosphere, the omission of polarization effects
may preclude a physically correct description of the forces driving
its conformational behavior. Even for small molecules the dipole
moment is known to vary significantly when they are transferred
from the gas to liquid phase. For example, an isolatedwatermolecule
has a dipole moment of 1.85 D [8], while the average molecular
dipole is 2.1 D in the water dimer, increasing in larger water clusters
[9]. In the condensed phase it reaches a value between 2.4 and 2.6
D, as suggested from classical molecular dynamics (MD) simulations
of the dielectric properties [10, 11], and 2.95 D, as obtained from
ab initio MD simulations [12-14] and from analysis of experimental
data [15, 16]. Additionally, our recent computational studies [17,
18], as well as studies of others [19], based on polarizable MD
simulations indicate the water dipole moment to be noticeably
perturbed in the vicinity of the charged groups of proteins.
Moreover, dipole moments of various chemical groups in proteins
itself, such as peptide backbone and side chains, were shown to
behave quite differently in polarizable and non-polarizable environ-
ments [18, 20]. In particular, MD simulations of a number of fully
solvated proteins utilizing our recently developed CHARMM Drude-
2013 polarizable force field produce time series for the dipole
moments of side chains of individual residues characterized by
significant variability and systematically higher values compared to
the additive results [18]. A similar trend was observed for the dipole
moments of nucleic acid bases from MD simulations of DNA [20].
These observations indicate that the variations of the electronic
structure do impact the dynamics of the system and the microscopic
forces dictating the structural and dynamical properties.